id stringlengths 9 10 | submitter stringlengths 1 64 ⌀ | authors stringlengths 4 20.7k | title stringlengths 4 246 | comments stringlengths 1 523 ⌀ | journal-ref stringlengths 4 404 ⌀ | doi stringlengths 11 153 ⌀ | report-no stringlengths 2 254 ⌀ | categories stringlengths 5 98 | license stringclasses 9 values | orig_abstract stringlengths 14 3.35k | versions listlengths 1 60 | update_date stringlengths 10 10 | authors_parsed listlengths 1 1.35k | abstract stringlengths 11 3.34k |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2312.13115 | Youjia Li | Youjia Li, Jianjun Shi, Zheng Zhang | A Novel Approach for Rapid Development Based on ChatGPT and Prompt
Engineering | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Code generation stands as a powerful technique in modern software
development, improving development efficiency, reducing errors, and fostering
standardization and consistency. Recently, ChatGPT has exhibited immense
potential in automatic code generation. However, existing researches on code
generation lack guidance for practical software development process. In this
study, we utilized ChatGPT to develop a web-based code generation platform
consisting of key components: User Interface, Prompt Builder and Backend
Service. Specifically, Prompt Builder dynamically generated comprehensive
prompts to enhance model generation performance. We conducted experiments on 2
datasets, evaluating the generated code through 8 widely used metrics.The
results demonstrate that (1) Our Prompt Builder is effective, resulting in a
65.06% improvement in EM, a 38.45% improvement in BLEU, a 15.70% improvement in
CodeBLEU, and a 50.64% improvement in Pass@1. (2) In real development
scenarios, 98.5% of test cases can be validated through manual validation,
highlighting the genuine assistance provided by the ChatGPT-based code
generation approach.
| [
{
"created": "Wed, 20 Dec 2023 15:36:13 GMT",
"version": "v1"
},
{
"created": "Thu, 21 Dec 2023 03:28:41 GMT",
"version": "v2"
}
] | 2023-12-22 | [
[
"Li",
"Youjia",
""
],
[
"Shi",
"Jianjun",
""
],
[
"Zhang",
"Zheng",
""
]
] | Code generation stands as a powerful technique in modern software development, improving development efficiency, reducing errors, and fostering standardization and consistency. Recently, ChatGPT has exhibited immense potential in automatic code generation. However, existing researches on code generation lack guidance for practical software development process. In this study, we utilized ChatGPT to develop a web-based code generation platform consisting of key components: User Interface, Prompt Builder and Backend Service. Specifically, Prompt Builder dynamically generated comprehensive prompts to enhance model generation performance. We conducted experiments on 2 datasets, evaluating the generated code through 8 widely used metrics.The results demonstrate that (1) Our Prompt Builder is effective, resulting in a 65.06% improvement in EM, a 38.45% improvement in BLEU, a 15.70% improvement in CodeBLEU, and a 50.64% improvement in Pass@1. (2) In real development scenarios, 98.5% of test cases can be validated through manual validation, highlighting the genuine assistance provided by the ChatGPT-based code generation approach. |
1705.01963 | Ray Li | Venkatesan Guruswami and Ray Li | Polynomial time decodable codes for the binary deletion channel | arXiv admin note: substantial text overlap with arXiv:1612.06335. The
published version of this paper incorrectly states the alphabet size in
Theorem 3.4. This version states the result correctly | IEEE Trans. Information Theory 65(4): 2171 - 2178 (2019) | null | null | cs.IT cs.DS math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the random deletion channel, each bit is deleted independently with
probability $p$. For the random deletion channel, the existence of codes of
rate $(1-p)/9$, and thus bounded away from $0$ for any $p < 1$, has been known.
We give an explicit construction with polynomial time encoding and deletion
correction algorithms with rate $c_0 (1-p)$ for an absolute constant $c_0 > 0$.
| [
{
"created": "Thu, 4 May 2017 18:18:07 GMT",
"version": "v1"
},
{
"created": "Wed, 26 Jul 2017 17:10:26 GMT",
"version": "v2"
},
{
"created": "Tue, 11 Jun 2019 23:12:58 GMT",
"version": "v3"
}
] | 2019-06-13 | [
[
"Guruswami",
"Venkatesan",
""
],
[
"Li",
"Ray",
""
]
] | In the random deletion channel, each bit is deleted independently with probability $p$. For the random deletion channel, the existence of codes of rate $(1-p)/9$, and thus bounded away from $0$ for any $p < 1$, has been known. We give an explicit construction with polynomial time encoding and deletion correction algorithms with rate $c_0 (1-p)$ for an absolute constant $c_0 > 0$. |
2311.07763 | Brian Barr PhD | Brian Barr, Noah Fatsi, Leif Hancox-Li, Peter Richter, Daniel Proano,
and Caleb Mok | The Disagreement Problem in Faithfulness Metrics | 6 pages (excluding refs and appendix) | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by-nc-sa/4.0/ | The field of explainable artificial intelligence (XAI) aims to explain how
black-box machine learning models work. Much of the work centers around the
holy grail of providing post-hoc feature attributions to any model
architecture. While the pace of innovation around novel methods has slowed
down, the question remains of how to choose a method, and how to make it fit
for purpose. Recently, efforts around benchmarking XAI methods have suggested
metrics for that purpose -- but there are many choices. That bounty of choice
still leaves an end user unclear on how to proceed. This paper focuses on
comparing metrics with the aim of measuring faithfulness of local explanations
on tabular classification problems -- and shows that the current metrics don't
agree; leaving users unsure how to choose the most faithful explanations.
| [
{
"created": "Mon, 13 Nov 2023 21:26:24 GMT",
"version": "v1"
}
] | 2023-11-15 | [
[
"Barr",
"Brian",
""
],
[
"Fatsi",
"Noah",
""
],
[
"Hancox-Li",
"Leif",
""
],
[
"Richter",
"Peter",
""
],
[
"Proano",
"Daniel",
""
],
[
"Mok",
"Caleb",
""
]
] | The field of explainable artificial intelligence (XAI) aims to explain how black-box machine learning models work. Much of the work centers around the holy grail of providing post-hoc feature attributions to any model architecture. While the pace of innovation around novel methods has slowed down, the question remains of how to choose a method, and how to make it fit for purpose. Recently, efforts around benchmarking XAI methods have suggested metrics for that purpose -- but there are many choices. That bounty of choice still leaves an end user unclear on how to proceed. This paper focuses on comparing metrics with the aim of measuring faithfulness of local explanations on tabular classification problems -- and shows that the current metrics don't agree; leaving users unsure how to choose the most faithful explanations. |
2301.11518 | Geng Zhao | Geng Zhao, Banghua Zhu, Jiantao Jiao, Michael I. Jordan | Online Learning in Stackelberg Games with an Omniscient Follower | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the problem of online learning in a two-player decentralized
cooperative Stackelberg game. In each round, the leader first takes an action,
followed by the follower who takes their action after observing the leader's
move. The goal of the leader is to learn to minimize the cumulative regret
based on the history of interactions. Differing from the traditional
formulation of repeated Stackelberg games, we assume the follower is
omniscient, with full knowledge of the true reward, and that they always
best-respond to the leader's actions. We analyze the sample complexity of
regret minimization in this repeated Stackelberg game. We show that depending
on the reward structure, the existence of the omniscient follower may change
the sample complexity drastically, from constant to exponential, even for
linear cooperative Stackelberg games. This poses unique challenges for the
learning process of the leader and the subsequent regret analysis.
| [
{
"created": "Fri, 27 Jan 2023 03:35:10 GMT",
"version": "v1"
},
{
"created": "Tue, 11 Apr 2023 20:27:37 GMT",
"version": "v2"
}
] | 2023-04-13 | [
[
"Zhao",
"Geng",
""
],
[
"Zhu",
"Banghua",
""
],
[
"Jiao",
"Jiantao",
""
],
[
"Jordan",
"Michael I.",
""
]
] | We study the problem of online learning in a two-player decentralized cooperative Stackelberg game. In each round, the leader first takes an action, followed by the follower who takes their action after observing the leader's move. The goal of the leader is to learn to minimize the cumulative regret based on the history of interactions. Differing from the traditional formulation of repeated Stackelberg games, we assume the follower is omniscient, with full knowledge of the true reward, and that they always best-respond to the leader's actions. We analyze the sample complexity of regret minimization in this repeated Stackelberg game. We show that depending on the reward structure, the existence of the omniscient follower may change the sample complexity drastically, from constant to exponential, even for linear cooperative Stackelberg games. This poses unique challenges for the learning process of the leader and the subsequent regret analysis. |
2205.13206 | Achille Felicetti | Franco Niccolucci, Achille Felicetti, Sorin Hermon | Populating the Digital Space for Cultural Heritage with Heritage Digital
Twins | Submitted to Data - An Open Access Journal from MDPI. 29 pages, 9
figures | Data 2022, 7(8), 105 | 10.3390/data7080105 | null | cs.DL cs.CY | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The present paper concerns the design of the semantic infrastructure of the
digital space for cultural heritage as envisaged by the European Commission in
its recent documents. Due to the complexity of the cultural heritage data and
of their intrinsic interrelationships, it is necessary to introduce a novel
ontology, yet compliant with existing standards and interoperable with previous
platforms used in this context, such as Europeana. The digital space
organization must be tailored to the methods and the theory of cultural
heritage, briefly summarized in the introduction. The new ontology is based on
the Digital Twin concept, i.e. the digital counterpart of cultural heritage
assets incorporating all the digital information pertaining to them. This
creates a Knowledge Base on the cultural heritage digital space. The paper
outlines the main features of the proposed Heritage Digital Twin ontology and
provides some examples of application. Future work will include completing the
ontology in all its details and testing it in other real cases and with the
various sectors of the cultural heritage community.
| [
{
"created": "Thu, 26 May 2022 07:49:27 GMT",
"version": "v1"
}
] | 2023-02-16 | [
[
"Niccolucci",
"Franco",
""
],
[
"Felicetti",
"Achille",
""
],
[
"Hermon",
"Sorin",
""
]
] | The present paper concerns the design of the semantic infrastructure of the digital space for cultural heritage as envisaged by the European Commission in its recent documents. Due to the complexity of the cultural heritage data and of their intrinsic interrelationships, it is necessary to introduce a novel ontology, yet compliant with existing standards and interoperable with previous platforms used in this context, such as Europeana. The digital space organization must be tailored to the methods and the theory of cultural heritage, briefly summarized in the introduction. The new ontology is based on the Digital Twin concept, i.e. the digital counterpart of cultural heritage assets incorporating all the digital information pertaining to them. This creates a Knowledge Base on the cultural heritage digital space. The paper outlines the main features of the proposed Heritage Digital Twin ontology and provides some examples of application. Future work will include completing the ontology in all its details and testing it in other real cases and with the various sectors of the cultural heritage community. |
2405.18751 | Jordi Armengol-Estap\'e | Jordi Armengol-Estap\'e, Vincent Michalski, Ramnath Kumar, Pierre-Luc
St-Charles, Doina Precup and Samira Ebrahimi Kahou | On the Limits of Multi-modal Meta-Learning with Auxiliary Task
Modulation Using Conditional Batch Normalization | null | null | null | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | Few-shot learning aims to learn representations that can tackle novel tasks
given a small number of examples. Recent studies show that cross-modal learning
can improve representations for few-shot classification. More specifically,
language is a rich modality that can be used to guide visual learning. In this
work, we experiment with a multi-modal architecture for few-shot learning that
consists of three components: a classifier, an auxiliary network, and a bridge
network. While the classifier performs the main classification task, the
auxiliary network learns to predict language representations from the same
input, and the bridge network transforms high-level features of the auxiliary
network into modulation parameters for layers of the few-shot classifier using
conditional batch normalization. The bridge should encourage a form of
lightweight semantic alignment between language and vision which could be
useful for the classifier. However, after evaluating the proposed approach on
two popular few-shot classification benchmarks we find that a) the improvements
do not reproduce across benchmarks, and b) when they do, the improvements are
due to the additional compute and parameters introduced by the bridge network.
We contribute insights and recommendations for future work in multi-modal
meta-learning, especially when using language representations.
| [
{
"created": "Wed, 29 May 2024 04:29:12 GMT",
"version": "v1"
},
{
"created": "Thu, 30 May 2024 14:13:05 GMT",
"version": "v2"
}
] | 2024-05-31 | [
[
"Armengol-Estapé",
"Jordi",
""
],
[
"Michalski",
"Vincent",
""
],
[
"Kumar",
"Ramnath",
""
],
[
"St-Charles",
"Pierre-Luc",
""
],
[
"Precup",
"Doina",
""
],
[
"Kahou",
"Samira Ebrahimi",
""
]
] | Few-shot learning aims to learn representations that can tackle novel tasks given a small number of examples. Recent studies show that cross-modal learning can improve representations for few-shot classification. More specifically, language is a rich modality that can be used to guide visual learning. In this work, we experiment with a multi-modal architecture for few-shot learning that consists of three components: a classifier, an auxiliary network, and a bridge network. While the classifier performs the main classification task, the auxiliary network learns to predict language representations from the same input, and the bridge network transforms high-level features of the auxiliary network into modulation parameters for layers of the few-shot classifier using conditional batch normalization. The bridge should encourage a form of lightweight semantic alignment between language and vision which could be useful for the classifier. However, after evaluating the proposed approach on two popular few-shot classification benchmarks we find that a) the improvements do not reproduce across benchmarks, and b) when they do, the improvements are due to the additional compute and parameters introduced by the bridge network. We contribute insights and recommendations for future work in multi-modal meta-learning, especially when using language representations. |
1609.03545 | Hayko Riemenschneider | Julien Weissenberg and Hayko Riemenschneider and Ralf Dragon and Luc
Van Gool | Dilemma First Search for Effortless Optimization of NP-Hard Problems | To be published at ICPR 2016 | null | null | null | cs.DS cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | To tackle the exponentiality associated with NP-hard problems, two paradigms
have been proposed. First, Branch & Bound, like Dynamic Programming, achieve
efficient exact inference but requires extensive information and analysis about
the problem at hand. Second, meta-heuristics are easier to implement but
comparatively inefficient. As a result, a number of problems have been left
unoptimized and plain greedy solutions are used. We introduce a theoretical
framework and propose a powerful yet simple search method called Dilemma First
Search (DFS). DFS exploits the decision heuristic needed for the greedy
solution for further optimization. DFS is useful when it is hard to design
efficient exact inference. We evaluate DFS on two problems: First, the Knapsack
problem, for which efficient algorithms exist, serves as a toy example. Second,
Decision Tree inference, where state-of-the-art algorithms rely on the greedy
or randomness-based solutions. We further show that decision trees benefit from
optimizations that are performed in a fraction of the iterations required by a
random-based search.
| [
{
"created": "Mon, 12 Sep 2016 19:36:02 GMT",
"version": "v1"
}
] | 2016-09-13 | [
[
"Weissenberg",
"Julien",
""
],
[
"Riemenschneider",
"Hayko",
""
],
[
"Dragon",
"Ralf",
""
],
[
"Van Gool",
"Luc",
""
]
] | To tackle the exponentiality associated with NP-hard problems, two paradigms have been proposed. First, Branch & Bound, like Dynamic Programming, achieve efficient exact inference but requires extensive information and analysis about the problem at hand. Second, meta-heuristics are easier to implement but comparatively inefficient. As a result, a number of problems have been left unoptimized and plain greedy solutions are used. We introduce a theoretical framework and propose a powerful yet simple search method called Dilemma First Search (DFS). DFS exploits the decision heuristic needed for the greedy solution for further optimization. DFS is useful when it is hard to design efficient exact inference. We evaluate DFS on two problems: First, the Knapsack problem, for which efficient algorithms exist, serves as a toy example. Second, Decision Tree inference, where state-of-the-art algorithms rely on the greedy or randomness-based solutions. We further show that decision trees benefit from optimizations that are performed in a fraction of the iterations required by a random-based search. |
2110.08797 | Gen Luo | Gen Luo, Yiyi Zhou, Xiaoshuai Sun, Yongjian Wu, Yue Gao, Rongrong Ji | Towards Language-guided Visual Recognition via Dynamic Convolutions | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we are committed to establishing an unified and end-to-end
multi-modal network via exploring the language-guided visual recognition. To
approach this target, we first propose a novel multi-modal convolution module
called Language-dependent Convolution (LaConv). Its convolution kernels are
dynamically generated based on natural language information, which can help
extract differentiated visual features for different multi-modal examples.
Based on the LaConv module, we further build the first fully language-driven
convolution network, termed as LaConvNet, which can unify the visual
recognition and multi-modal reasoning in one forward structure. To validate
LaConv and LaConvNet, we conduct extensive experiments on four benchmark
datasets of two vision-and-language tasks, i.e., visual question answering
(VQA) and referring expression comprehension (REC). The experimental results
not only shows the performance gains of LaConv compared to the existing
multi-modal modules, but also witness the merits of LaConvNet as an unified
network, including compact network, high generalization ability and excellent
performance, e.g., +4.7% on RefCOCO+.
| [
{
"created": "Sun, 17 Oct 2021 11:29:13 GMT",
"version": "v1"
},
{
"created": "Thu, 14 Sep 2023 13:37:38 GMT",
"version": "v2"
}
] | 2023-09-15 | [
[
"Luo",
"Gen",
""
],
[
"Zhou",
"Yiyi",
""
],
[
"Sun",
"Xiaoshuai",
""
],
[
"Wu",
"Yongjian",
""
],
[
"Gao",
"Yue",
""
],
[
"Ji",
"Rongrong",
""
]
] | In this paper, we are committed to establishing an unified and end-to-end multi-modal network via exploring the language-guided visual recognition. To approach this target, we first propose a novel multi-modal convolution module called Language-dependent Convolution (LaConv). Its convolution kernels are dynamically generated based on natural language information, which can help extract differentiated visual features for different multi-modal examples. Based on the LaConv module, we further build the first fully language-driven convolution network, termed as LaConvNet, which can unify the visual recognition and multi-modal reasoning in one forward structure. To validate LaConv and LaConvNet, we conduct extensive experiments on four benchmark datasets of two vision-and-language tasks, i.e., visual question answering (VQA) and referring expression comprehension (REC). The experimental results not only shows the performance gains of LaConv compared to the existing multi-modal modules, but also witness the merits of LaConvNet as an unified network, including compact network, high generalization ability and excellent performance, e.g., +4.7% on RefCOCO+. |
1611.03971 | Tayfun Tuna | Tayfun Tuna and Esra Akbas and Ahmet Aksoy and Muhammed Abdullah
Canbaz and Umit Karabiyik and Bilal Gonen and Ramazan Aygun | User characterization for online social networks | null | Soc. Netw. Anal. Min. (2016) 6: 104. doi:10.1007/s13278-016-0412-3 | 10.1007/s13278-016-0412-3 | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Online social network analysis has attracted great attention with a vast
number of users sharing information and availability of APIs that help to crawl
online social network data. In this paper, we study the research studies that
are helpful for user characterization as online users may not always reveal
their true identity or attributes. We especially focused on user attribute
determination such as gender, age, etc.; user behavior analysis such as motives
for deception; mental models that are indicators of user behavior; user
categorization such as bots vs. humans; and entity matching on different social
networks. We believe our summary of analysis of user characterization will
provide important insights to researchers and better services to online users.
| [
{
"created": "Sat, 12 Nov 2016 08:30:18 GMT",
"version": "v1"
},
{
"created": "Tue, 27 Dec 2016 00:11:13 GMT",
"version": "v2"
}
] | 2016-12-28 | [
[
"Tuna",
"Tayfun",
""
],
[
"Akbas",
"Esra",
""
],
[
"Aksoy",
"Ahmet",
""
],
[
"Canbaz",
"Muhammed Abdullah",
""
],
[
"Karabiyik",
"Umit",
""
],
[
"Gonen",
"Bilal",
""
],
[
"Aygun",
"Ramazan",
""
]
] | Online social network analysis has attracted great attention with a vast number of users sharing information and availability of APIs that help to crawl online social network data. In this paper, we study the research studies that are helpful for user characterization as online users may not always reveal their true identity or attributes. We especially focused on user attribute determination such as gender, age, etc.; user behavior analysis such as motives for deception; mental models that are indicators of user behavior; user categorization such as bots vs. humans; and entity matching on different social networks. We believe our summary of analysis of user characterization will provide important insights to researchers and better services to online users. |
2307.14208 | Tanapol Kosolwattana | Tanapol Kosolwattana, Huazheng Wang, Ying Lin | Online Modeling and Monitoring of Dependent Processes under Resource
Constraints | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Adaptive monitoring of a large population of dynamic processes is critical
for the timely detection of abnormal events under limited resources in many
healthcare and engineering systems. Examples include the risk-based disease
screening and condition-based process monitoring. However, existing adaptive
monitoring models either ignore the dependency among processes or overlook the
uncertainty in process modeling. To design an optimal monitoring strategy that
accurately monitors the processes with poor health conditions and actively
collects information for uncertainty reduction, a novel online collaborative
learning method is proposed in this study. The proposed method designs a
collaborative learning-based upper confidence bound (CL-UCB) algorithm to
optimally balance the exploitation and exploration of dependent processes under
limited resources. Efficiency of the proposed method is demonstrated through
theoretical analysis, simulation studies and an empirical study of adaptive
cognitive monitoring in Alzheimer's disease.
| [
{
"created": "Wed, 26 Jul 2023 14:14:38 GMT",
"version": "v1"
},
{
"created": "Sat, 21 Oct 2023 23:14:34 GMT",
"version": "v2"
}
] | 2023-10-24 | [
[
"Kosolwattana",
"Tanapol",
""
],
[
"Wang",
"Huazheng",
""
],
[
"Lin",
"Ying",
""
]
] | Adaptive monitoring of a large population of dynamic processes is critical for the timely detection of abnormal events under limited resources in many healthcare and engineering systems. Examples include the risk-based disease screening and condition-based process monitoring. However, existing adaptive monitoring models either ignore the dependency among processes or overlook the uncertainty in process modeling. To design an optimal monitoring strategy that accurately monitors the processes with poor health conditions and actively collects information for uncertainty reduction, a novel online collaborative learning method is proposed in this study. The proposed method designs a collaborative learning-based upper confidence bound (CL-UCB) algorithm to optimally balance the exploitation and exploration of dependent processes under limited resources. Efficiency of the proposed method is demonstrated through theoretical analysis, simulation studies and an empirical study of adaptive cognitive monitoring in Alzheimer's disease. |
2303.05807 | Ziteng Cui | Ziteng Cui, Lin Gu, Xiao Sun, Xianzheng Ma, Yu Qiao, Tatsuya Harada | Aleth-NeRF: Low-light Condition View Synthesis with Concealing Fields | website page: https://cuiziteng.github.io/Aleth_NeRF_web/, refer to
new version: arXiv:2312.09093 | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Common capture low-light scenes are challenging for most computer vision
techniques, including Neural Radiance Fields (NeRF). Vanilla NeRF is
viewer-centred simplifies the rendering process only as light emission from 3D
locations in the viewing direction, thus failing to model the low-illumination
induced darkness. Inspired by the emission theory of ancient Greeks that visual
perception is accomplished by rays casting from eyes, we make slight
modifications on vanilla NeRF to train on multiple views of low-light scenes,
we can thus render out the well-lit scene in an unsupervised manner. We
introduce a surrogate concept, Concealing Fields, that reduces the transport of
light during the volume rendering stage. Specifically, our proposed method,
Aleth-NeRF, directly learns from the dark image to understand volumetric object
representation and concealing field under priors. By simply eliminating
Concealing Fields, we can render a single or multi-view well-lit image(s) and
gain superior performance over other 2D low-light enhancement methods.
Additionally, we collect the first paired LOw-light and normal-light Multi-view
(LOM) datasets for future research. This version is invalid, please refer to
our new AAAI version: arXiv:2312.09093
| [
{
"created": "Fri, 10 Mar 2023 09:28:09 GMT",
"version": "v1"
},
{
"created": "Sat, 30 Dec 2023 02:42:12 GMT",
"version": "v2"
}
] | 2024-01-02 | [
[
"Cui",
"Ziteng",
""
],
[
"Gu",
"Lin",
""
],
[
"Sun",
"Xiao",
""
],
[
"Ma",
"Xianzheng",
""
],
[
"Qiao",
"Yu",
""
],
[
"Harada",
"Tatsuya",
""
]
] | Common capture low-light scenes are challenging for most computer vision techniques, including Neural Radiance Fields (NeRF). Vanilla NeRF is viewer-centred simplifies the rendering process only as light emission from 3D locations in the viewing direction, thus failing to model the low-illumination induced darkness. Inspired by the emission theory of ancient Greeks that visual perception is accomplished by rays casting from eyes, we make slight modifications on vanilla NeRF to train on multiple views of low-light scenes, we can thus render out the well-lit scene in an unsupervised manner. We introduce a surrogate concept, Concealing Fields, that reduces the transport of light during the volume rendering stage. Specifically, our proposed method, Aleth-NeRF, directly learns from the dark image to understand volumetric object representation and concealing field under priors. By simply eliminating Concealing Fields, we can render a single or multi-view well-lit image(s) and gain superior performance over other 2D low-light enhancement methods. Additionally, we collect the first paired LOw-light and normal-light Multi-view (LOM) datasets for future research. This version is invalid, please refer to our new AAAI version: arXiv:2312.09093 |
1910.08185 | Wail Alkowaileet | Wail Y. Alkowaileet, Sattam Alsubaiee and Michael J. Carey | An LSM-based Tuple Compaction Framework for Apache AsterixDB (Extended
Version) | 18 pages, 28 figures, to appear in VLDB 2020 | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Document database systems store self-describing semi-structured records, such
as JSON, "as-is" without requiring the users to pre-define a schema. This
provides users with the flexibility to change the structure of incoming records
without worrying about taking the system offline or hindering the performance
of currently running queries. However, the flexibility of such systems does not
free. The large amount of redundancy in the records can introduce an
unnecessary storage overhead and impact query performance.
Our focus in this paper is to address the storage overhead issue by
introducing a tuple compactor framework that infers and extracts the schema
from self-describing semi-structured records during the data ingestion. As many
prominent document stores, such as MongoDB and Couchbase, adopt Log Structured
Merge (LSM) trees in their storage engines, our framework exploits LSM
lifecycle events to piggyback the schema inference and extraction operations.
We have implemented and empirically evaluated our approach to measure its
impact on storage, data ingestion, and query performance in the context of
Apache AsterixDB.
| [
{
"created": "Thu, 17 Oct 2019 22:13:40 GMT",
"version": "v1"
},
{
"created": "Mon, 11 May 2020 05:23:31 GMT",
"version": "v2"
}
] | 2020-05-12 | [
[
"Alkowaileet",
"Wail Y.",
""
],
[
"Alsubaiee",
"Sattam",
""
],
[
"Carey",
"Michael J.",
""
]
] | Document database systems store self-describing semi-structured records, such as JSON, "as-is" without requiring the users to pre-define a schema. This provides users with the flexibility to change the structure of incoming records without worrying about taking the system offline or hindering the performance of currently running queries. However, the flexibility of such systems does not free. The large amount of redundancy in the records can introduce an unnecessary storage overhead and impact query performance. Our focus in this paper is to address the storage overhead issue by introducing a tuple compactor framework that infers and extracts the schema from self-describing semi-structured records during the data ingestion. As many prominent document stores, such as MongoDB and Couchbase, adopt Log Structured Merge (LSM) trees in their storage engines, our framework exploits LSM lifecycle events to piggyback the schema inference and extraction operations. We have implemented and empirically evaluated our approach to measure its impact on storage, data ingestion, and query performance in the context of Apache AsterixDB. |
1302.4767 | Nicola Laurenti | Francesco Renna, Nicola Laurenti, Stefano Tomasin, Marco Baldi, Nicola
Maturo, Marco Bianchi, Franco Chiaraluce and Matthieu Bloch | Low-power Secret-key Agreement over OFDM | 9 pages, 4 figures; this is the authors prepared version of the paper
with the same name accepted for HotWiSec 2013, the Second ACM Workshop on Hot
Topics on Wireless Network Security and Privacy, Budapest, Hungary 17-19
April 2013 | null | null | null | cs.IT cs.CR math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Information-theoretic secret-key agreement is perhaps the most practically
feasible mechanism that provides unconditional security at the physical layer
to date. In this paper, we consider the problem of secret-key agreement by
sharing randomness at low power over an orthogonal frequency division
multiplexing (OFDM) link, in the presence of an eavesdropper. The low power
assumption greatly simplifies the design of the randomness sharing scheme, even
in a fading channel scenario. We assess the performance of the proposed system
in terms of secrecy key rate and show that a practical approach to key sharing
is obtained by using low-density parity check (LDPC) codes for information
reconciliation. Numerical results confirm the merits of the proposed approach
as a feasible and practical solution. Moreover, the outage formulation allows
to implement secret-key agreement even when only statistical knowledge of the
eavesdropper channel is available.
| [
{
"created": "Tue, 19 Feb 2013 22:19:12 GMT",
"version": "v1"
}
] | 2013-02-21 | [
[
"Renna",
"Francesco",
""
],
[
"Laurenti",
"Nicola",
""
],
[
"Tomasin",
"Stefano",
""
],
[
"Baldi",
"Marco",
""
],
[
"Maturo",
"Nicola",
""
],
[
"Bianchi",
"Marco",
""
],
[
"Chiaraluce",
"Franco",
""
],
[
"Bloch",
"Matthieu",
""
]
] | Information-theoretic secret-key agreement is perhaps the most practically feasible mechanism that provides unconditional security at the physical layer to date. In this paper, we consider the problem of secret-key agreement by sharing randomness at low power over an orthogonal frequency division multiplexing (OFDM) link, in the presence of an eavesdropper. The low power assumption greatly simplifies the design of the randomness sharing scheme, even in a fading channel scenario. We assess the performance of the proposed system in terms of secrecy key rate and show that a practical approach to key sharing is obtained by using low-density parity check (LDPC) codes for information reconciliation. Numerical results confirm the merits of the proposed approach as a feasible and practical solution. Moreover, the outage formulation allows to implement secret-key agreement even when only statistical knowledge of the eavesdropper channel is available. |
2407.01976 | Jinghui Lu | Jinghui Lu, Haiyang Yu, Yanjie Wang, Yongjie Ye, Jingqun Tang, Ziwei
Yang, Binghong Wu, Qi Liu, Hao Feng, Han Wang, Hao Liu, Can Huang | A Bounding Box is Worth One Token: Interleaving Layout and Text in a
Large Language Model for Document Understanding | null | null | null | null | cs.CL cs.AI cs.MM | http://creativecommons.org/publicdomain/zero/1.0/ | Recently, many studies have demonstrated that exclusively incorporating
OCR-derived text and spatial layouts with large language models (LLMs) can be
highly effective for document understanding tasks. However, existing methods
that integrate spatial layouts with text have limitations, such as producing
overly long text sequences or failing to fully leverage the autoregressive
traits of LLMs. In this work, we introduce Interleaving Layout and Text in a
Large Language Model (LayTextLLM)} for document understanding. In particular,
LayTextLLM projects each bounding box to a single embedding and interleaves it
with text, efficiently avoiding long sequence issues while leveraging
autoregressive traits of LLMs. LayTextLLM not only streamlines the interaction
of layout and textual data but also shows enhanced performance in Key
Information Extraction (KIE) and Visual Question Answering (VQA). Comprehensive
benchmark evaluations reveal significant improvements, with a 27.2% increase on
KIE tasks and 12.0% on VQA tasks compared to previous state-of-the-art document
understanding MLLMs, as well as a 15.1% improvement over other SOTA OCR-based
LLMs on KIE tasks.
| [
{
"created": "Tue, 2 Jul 2024 06:29:05 GMT",
"version": "v1"
},
{
"created": "Wed, 24 Jul 2024 11:45:48 GMT",
"version": "v2"
}
] | 2024-07-25 | [
[
"Lu",
"Jinghui",
""
],
[
"Yu",
"Haiyang",
""
],
[
"Wang",
"Yanjie",
""
],
[
"Ye",
"Yongjie",
""
],
[
"Tang",
"Jingqun",
""
],
[
"Yang",
"Ziwei",
""
],
[
"Wu",
"Binghong",
""
],
[
"Liu",
"Qi",
""
],
[
"Feng",
"Hao",
""
],
[
"Wang",
"Han",
""
],
[
"Liu",
"Hao",
""
],
[
"Huang",
"Can",
""
]
] | Recently, many studies have demonstrated that exclusively incorporating OCR-derived text and spatial layouts with large language models (LLMs) can be highly effective for document understanding tasks. However, existing methods that integrate spatial layouts with text have limitations, such as producing overly long text sequences or failing to fully leverage the autoregressive traits of LLMs. In this work, we introduce Interleaving Layout and Text in a Large Language Model (LayTextLLM)} for document understanding. In particular, LayTextLLM projects each bounding box to a single embedding and interleaves it with text, efficiently avoiding long sequence issues while leveraging autoregressive traits of LLMs. LayTextLLM not only streamlines the interaction of layout and textual data but also shows enhanced performance in Key Information Extraction (KIE) and Visual Question Answering (VQA). Comprehensive benchmark evaluations reveal significant improvements, with a 27.2% increase on KIE tasks and 12.0% on VQA tasks compared to previous state-of-the-art document understanding MLLMs, as well as a 15.1% improvement over other SOTA OCR-based LLMs on KIE tasks. |
2007.05216 | Sajan Kedia | Sajan Kedia, Samyak Jain, Abhishek Sharma | Price Optimization in Fashion E-commerce | 8 pages, 6 figures, AI for fashion supply chain Conference | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the rapid growth in the fashion e-commerce industry, it is becoming
extremely challenging for the E-tailers to set an optimal price point for all
the products on the platform. By establishing an optimal price point, they can
maximize overall revenue and profit for the platform. In this paper, we propose
a novel machine learning and optimization technique to find the optimal price
point at an individual product level. It comprises three major components.
Firstly, we use a demand prediction model to predict the next day demand for
each product at a certain discount percentage. Next step, we use the concept of
price elasticity of demand to get the multiple demand values by varying the
discount percentage. Thus we obtain multiple price demand pairs for each
product and we have to choose one of them for the live platform. Typically
fashion e-commerce has millions of products, so there can be many permutations.
Each permutation will assign a unique price point for all the products, which
will sum up to a unique revenue number. To choose the best permutation which
gives maximum revenue, a linear programming optimization technique is used. We
have deployed the above methods in the live production environment and
conducted several AB tests. According to the AB test result, our model is
improving the revenue by 1 percent and gross margin by 0.81 percent.
| [
{
"created": "Fri, 10 Jul 2020 07:40:28 GMT",
"version": "v1"
},
{
"created": "Mon, 24 Aug 2020 10:18:53 GMT",
"version": "v2"
}
] | 2020-08-25 | [
[
"Kedia",
"Sajan",
""
],
[
"Jain",
"Samyak",
""
],
[
"Sharma",
"Abhishek",
""
]
] | With the rapid growth in the fashion e-commerce industry, it is becoming extremely challenging for the E-tailers to set an optimal price point for all the products on the platform. By establishing an optimal price point, they can maximize overall revenue and profit for the platform. In this paper, we propose a novel machine learning and optimization technique to find the optimal price point at an individual product level. It comprises three major components. Firstly, we use a demand prediction model to predict the next day demand for each product at a certain discount percentage. Next step, we use the concept of price elasticity of demand to get the multiple demand values by varying the discount percentage. Thus we obtain multiple price demand pairs for each product and we have to choose one of them for the live platform. Typically fashion e-commerce has millions of products, so there can be many permutations. Each permutation will assign a unique price point for all the products, which will sum up to a unique revenue number. To choose the best permutation which gives maximum revenue, a linear programming optimization technique is used. We have deployed the above methods in the live production environment and conducted several AB tests. According to the AB test result, our model is improving the revenue by 1 percent and gross margin by 0.81 percent. |
1901.07634 | Lam Nguyen | Lam M. Nguyen, Phuong Ha Nguyen, Dzung T. Phan, Jayant R. Kalagnanam,
Marten van Dijk | DTN: A Learning Rate Scheme with Convergence Rate of $\mathcal{O}(1/t)$
for SGD | This paper has inconsistent results, i.e., we made some failed claims
because we did some mistakes for using the test criterion for a series | null | null | null | cs.LG math.OC stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper has some inconsistent results, i.e., we made some failed claims
because we did some mistakes for using the test criterion for a series.
Precisely, our claims on the convergence rate of $\mathcal{O}(1/t)$ of SGD
presented in Theorem 1, Corollary 1, Theorem 2 and Corollary 2 are wrongly
derived because they are based on Lemma 5. In Lemma 5, we do not correctly use
the test criterion for a series. Hence, the result of Lemma 5 is not valid. We
would like to thank the community for pointing out this mistake!
| [
{
"created": "Tue, 22 Jan 2019 22:40:31 GMT",
"version": "v1"
},
{
"created": "Mon, 28 Jan 2019 21:55:19 GMT",
"version": "v2"
},
{
"created": "Thu, 28 Feb 2019 02:01:20 GMT",
"version": "v3"
}
] | 2019-03-01 | [
[
"Nguyen",
"Lam M.",
""
],
[
"Nguyen",
"Phuong Ha",
""
],
[
"Phan",
"Dzung T.",
""
],
[
"Kalagnanam",
"Jayant R.",
""
],
[
"van Dijk",
"Marten",
""
]
] | This paper has some inconsistent results, i.e., we made some failed claims because we did some mistakes for using the test criterion for a series. Precisely, our claims on the convergence rate of $\mathcal{O}(1/t)$ of SGD presented in Theorem 1, Corollary 1, Theorem 2 and Corollary 2 are wrongly derived because they are based on Lemma 5. In Lemma 5, we do not correctly use the test criterion for a series. Hence, the result of Lemma 5 is not valid. We would like to thank the community for pointing out this mistake! |
2005.12712 | Benedikt Kleinmeier | Benedikt Kleinmeier, Gerta K\"oster, John Drury | Agent-Based Simulation of Collective Cooperation: From Experiment to
Model | 16 pages, 19 figures, 5 tables, 4 listings, interdisciplinary work
between computer science and psychology | Journal of the Royal Society Interface (October 2020, Volume 17,
Issue 171) | 10.1098/rsif.2020.0396 | null | cs.MA cs.CY | http://creativecommons.org/licenses/by/4.0/ | Simulation models of pedestrian dynamics have become an invaluable tool for
evacuation planning. Typically crowds are assumed to stream unidirectionally
towards a safe area. Simulated agents avoid collisions through mechanisms that
belong to each individual, such as being repelled from each other by imaginary
forces. But classic locomotion models fail when collective cooperation is
called for, notably when an agent, say a first-aid attendant, needs to forge a
path through a densely packed group. We present a controlled experiment to
observe what happens when humans pass through a dense static crowd. We
formulate and test hypothesis on salient phenomena. We discuss our observations
in a psychological framework. We derive a model that incorporates: agents'
perception and cognitive processing of a situation that needs cooperation;
selection from a portfolio of behaviours, such as being cooperative; and a
suitable action, such as swapping places. Agents' ability to successfully get
through a dense crowd emerges as an effect of the psychological model.
| [
{
"created": "Tue, 26 May 2020 13:29:08 GMT",
"version": "v1"
},
{
"created": "Wed, 7 Oct 2020 09:40:47 GMT",
"version": "v2"
}
] | 2020-10-08 | [
[
"Kleinmeier",
"Benedikt",
""
],
[
"Köster",
"Gerta",
""
],
[
"Drury",
"John",
""
]
] | Simulation models of pedestrian dynamics have become an invaluable tool for evacuation planning. Typically crowds are assumed to stream unidirectionally towards a safe area. Simulated agents avoid collisions through mechanisms that belong to each individual, such as being repelled from each other by imaginary forces. But classic locomotion models fail when collective cooperation is called for, notably when an agent, say a first-aid attendant, needs to forge a path through a densely packed group. We present a controlled experiment to observe what happens when humans pass through a dense static crowd. We formulate and test hypothesis on salient phenomena. We discuss our observations in a psychological framework. We derive a model that incorporates: agents' perception and cognitive processing of a situation that needs cooperation; selection from a portfolio of behaviours, such as being cooperative; and a suitable action, such as swapping places. Agents' ability to successfully get through a dense crowd emerges as an effect of the psychological model. |
2308.03019 | Pratapa Redy Lankireddy | Naveenkumar Vodnala (VNR Vignana Jyothi Institute of Engineering and
Technology), Pratap Reddy Lankireddy (Jawaharlal Nehru Technological
University Hyderabad), Padmasai Yarlagadda (VNR Vignana Jyothi Institute of
Engineering and Technology) | Characterization of cough sounds using statistical analysis | 19 pages, 8 figures, paper submitted to journal Biomedical Signal
Processing and Control which is under review | null | null | null | cs.SD eess.AS eess.SP | http://creativecommons.org/licenses/by-sa/4.0/ | Cough is a primary symptom of most respiratory diseases, and changes in cough
characteristics provide valuable information for diagnosing respiratory
diseases. The characterization of cough sounds still lacks concrete evidence,
which makes it difficult to accurately distinguish between different types of
coughs and other sounds. The objective of this research work is to characterize
cough sounds with voiced content and cough sounds without voiced content.
Further, the cough sound characteristics are compared with the characteristics
of speech. The proposed method to achieve this goal utilized spectral roll-off,
spectral entropy, spectral flatness, spectral flux, zero crossing rate,
spectral centroid, and spectral bandwidth attributes which describe the cough
sounds related to the respiratory system, glottal information, and voice model.
These attributes are then subjected to statistical analysis using the measures
of minimum, maximum, mean, median, and standard deviation. The experimental
results show that the mean and frequency distribution of spectral roll-off,
spectral centroid, and spectral bandwidth are found to be higher for cough
sounds than for speech signals. Spectral flatness levels in cough sounds will
rise to 0.22, whereas spectral flux varies between 0.3 and 0.6. The Zero
Crossing Rate (ZCR) of most frames of cough sounds is between 0.05 and 0.4.
These attributes contribute significant information while characterizing cough
sounds.
| [
{
"created": "Sun, 6 Aug 2023 04:26:52 GMT",
"version": "v1"
}
] | 2023-08-08 | [
[
"Vodnala",
"Naveenkumar",
"",
"VNR Vignana Jyothi Institute of Engineering and\n Technology"
],
[
"Lankireddy",
"Pratap Reddy",
"",
"Jawaharlal Nehru Technological\n University Hyderabad"
],
[
"Yarlagadda",
"Padmasai",
"",
"VNR Vignana Jyothi Institute of\n Engineering and Technology"
]
] | Cough is a primary symptom of most respiratory diseases, and changes in cough characteristics provide valuable information for diagnosing respiratory diseases. The characterization of cough sounds still lacks concrete evidence, which makes it difficult to accurately distinguish between different types of coughs and other sounds. The objective of this research work is to characterize cough sounds with voiced content and cough sounds without voiced content. Further, the cough sound characteristics are compared with the characteristics of speech. The proposed method to achieve this goal utilized spectral roll-off, spectral entropy, spectral flatness, spectral flux, zero crossing rate, spectral centroid, and spectral bandwidth attributes which describe the cough sounds related to the respiratory system, glottal information, and voice model. These attributes are then subjected to statistical analysis using the measures of minimum, maximum, mean, median, and standard deviation. The experimental results show that the mean and frequency distribution of spectral roll-off, spectral centroid, and spectral bandwidth are found to be higher for cough sounds than for speech signals. Spectral flatness levels in cough sounds will rise to 0.22, whereas spectral flux varies between 0.3 and 0.6. The Zero Crossing Rate (ZCR) of most frames of cough sounds is between 0.05 and 0.4. These attributes contribute significant information while characterizing cough sounds. |
1407.4668 | Albrecht Zimmermann | Albrecht Zimmermann | A feature construction framework based on outlier detection and
discriminative pattern mining | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | No matter the expressive power and sophistication of supervised learning
algorithms, their effectiveness is restricted by the features describing the
data. This is not a new insight in ML and many methods for feature selection,
transformation, and construction have been developed. But while this is
on-going for general techniques for feature selection and transformation, i.e.
dimensionality reduction, work on feature construction, i.e. enriching the
data, is by now mainly the domain of image, particularly character,
recognition, and NLP.
In this work, we propose a new general framework for feature construction.
The need for feature construction in a data set is indicated by class outliers
and discriminative pattern mining used to derive features on their
k-neighborhoods. We instantiate the framework with LOF and C4.5-Rules, and
evaluate the usefulness of the derived features on a diverse collection of UCI
data sets. The derived features are more often useful than ones derived by
DC-Fringe, and our approach is much less likely to overfit. But while a weak
learner, Naive Bayes, benefits strongly from the feature construction, the
effect is less pronounced for C4.5, and almost vanishes for an SVM leaner.
Keywords: feature construction, classification, outlier detection
| [
{
"created": "Thu, 17 Jul 2014 13:51:55 GMT",
"version": "v1"
}
] | 2014-07-18 | [
[
"Zimmermann",
"Albrecht",
""
]
] | No matter the expressive power and sophistication of supervised learning algorithms, their effectiveness is restricted by the features describing the data. This is not a new insight in ML and many methods for feature selection, transformation, and construction have been developed. But while this is on-going for general techniques for feature selection and transformation, i.e. dimensionality reduction, work on feature construction, i.e. enriching the data, is by now mainly the domain of image, particularly character, recognition, and NLP. In this work, we propose a new general framework for feature construction. The need for feature construction in a data set is indicated by class outliers and discriminative pattern mining used to derive features on their k-neighborhoods. We instantiate the framework with LOF and C4.5-Rules, and evaluate the usefulness of the derived features on a diverse collection of UCI data sets. The derived features are more often useful than ones derived by DC-Fringe, and our approach is much less likely to overfit. But while a weak learner, Naive Bayes, benefits strongly from the feature construction, the effect is less pronounced for C4.5, and almost vanishes for an SVM leaner. Keywords: feature construction, classification, outlier detection |
1711.04022 | Hamid Eghbal-zadeh | Hamid Eghbal-zadeh, Matthias Dorfer and Gerhard Widmer | Deep Within-Class Covariance Analysis for Robust Audio Representation
Learning | 11 pages, 3 tables, 4 figures | null | null | null | cs.LG cs.AI cs.SD eess.AS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Convolutional Neural Networks (CNNs) can learn effective features, though
have been shown to suffer from a performance drop when the distribution of the
data changes from training to test data. In this paper we analyze the internal
representations of CNNs and observe that the representations of unseen data in
each class, spread more (with higher variance) in the embedding space of the
CNN compared to representations of the training data. More importantly, this
difference is more extreme if the unseen data comes from a shifted
distribution. Based on this observation, we objectively evaluate the degree of
representation's variance in each class via eigenvalue decomposition on the
within-class covariance of the internal representations of CNNs and observe the
same behaviour. This can be problematic as larger variances might lead to
mis-classification if the sample crosses the decision boundary of its class. We
apply nearest neighbor classification on the representations and empirically
show that the embeddings with the high variance actually have significantly
worse KNN classification performances, although this could not be foreseen from
their end-to-end classification results. To tackle this problem, we propose
Deep Within-Class Covariance Analysis (DWCCA), a deep neural network layer that
significantly reduces the within-class covariance of a DNN's representation,
improving performance on unseen test data from a shifted distribution. We
empirically evaluate DWCCA on two datasets for Acoustic Scene Classification
(DCASE2016 and DCASE2017). We demonstrate that not only does DWCCA
significantly improve the network's internal representation, it also increases
the end-to-end classification accuracy, especially when the test set exhibits a
distribution shift. By adding DWCCA to a VGG network, we achieve around 6
percentage points improvement in the case of a distribution mismatch.
| [
{
"created": "Fri, 10 Nov 2017 21:39:12 GMT",
"version": "v1"
},
{
"created": "Fri, 30 Nov 2018 09:48:48 GMT",
"version": "v2"
}
] | 2018-12-03 | [
[
"Eghbal-zadeh",
"Hamid",
""
],
[
"Dorfer",
"Matthias",
""
],
[
"Widmer",
"Gerhard",
""
]
] | Convolutional Neural Networks (CNNs) can learn effective features, though have been shown to suffer from a performance drop when the distribution of the data changes from training to test data. In this paper we analyze the internal representations of CNNs and observe that the representations of unseen data in each class, spread more (with higher variance) in the embedding space of the CNN compared to representations of the training data. More importantly, this difference is more extreme if the unseen data comes from a shifted distribution. Based on this observation, we objectively evaluate the degree of representation's variance in each class via eigenvalue decomposition on the within-class covariance of the internal representations of CNNs and observe the same behaviour. This can be problematic as larger variances might lead to mis-classification if the sample crosses the decision boundary of its class. We apply nearest neighbor classification on the representations and empirically show that the embeddings with the high variance actually have significantly worse KNN classification performances, although this could not be foreseen from their end-to-end classification results. To tackle this problem, we propose Deep Within-Class Covariance Analysis (DWCCA), a deep neural network layer that significantly reduces the within-class covariance of a DNN's representation, improving performance on unseen test data from a shifted distribution. We empirically evaluate DWCCA on two datasets for Acoustic Scene Classification (DCASE2016 and DCASE2017). We demonstrate that not only does DWCCA significantly improve the network's internal representation, it also increases the end-to-end classification accuracy, especially when the test set exhibits a distribution shift. By adding DWCCA to a VGG network, we achieve around 6 percentage points improvement in the case of a distribution mismatch. |
2303.10567 | Jinyeong Jeong | Jinyeong Jeong, Min Jun Kim | Passivity-based Decentralized Control for Collaborative Grasping of
Under-Actuated Aerial Manipulators | IEEE International Conference on Robotics and Automation (ICRA) 2023 | null | 10.1109/ICRA48891.2023.10160334 | null | cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper proposes a decentralized passive impedance control scheme for
collaborative grasping using under-actuated aerial manipulators (AMs). The AM
system is formulated, using a proper coordinate transformation, as an
inertially decoupled dynamics with which a passivity-based control design is
conducted. Since the interaction for grasping can be interpreted as a feedback
interconnection of passive systems, an arbitrary number of AMs can be modularly
combined, leading to a decentralized control scheme. Another interesting
consequence of the passivity property is that the AMs automatically converge to
a certain configuration to accomplish the grasping. Collaborative grasping
using 10 AMs is presented in simulation.
| [
{
"created": "Sun, 19 Mar 2023 05:04:50 GMT",
"version": "v1"
}
] | 2024-01-12 | [
[
"Jeong",
"Jinyeong",
""
],
[
"Kim",
"Min Jun",
""
]
] | This paper proposes a decentralized passive impedance control scheme for collaborative grasping using under-actuated aerial manipulators (AMs). The AM system is formulated, using a proper coordinate transformation, as an inertially decoupled dynamics with which a passivity-based control design is conducted. Since the interaction for grasping can be interpreted as a feedback interconnection of passive systems, an arbitrary number of AMs can be modularly combined, leading to a decentralized control scheme. Another interesting consequence of the passivity property is that the AMs automatically converge to a certain configuration to accomplish the grasping. Collaborative grasping using 10 AMs is presented in simulation. |
1904.02074 | Qinbing Fu | Qinbing Fu, Nicola Bellotto, Huatian Wang, F. Claire Rind, Hongxin
Wang, Shigang Yue | A Visual Neural Network for Robust Collision Perception in Vehicle
Driving Scenarios | 12 pages, 7 figures, conference, springer format | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This research addresses the challenging problem of visual collision detection
in very complex and dynamic real physical scenes, specifically, the vehicle
driving scenarios. This research takes inspiration from a large-field looming
sensitive neuron, i.e., the lobula giant movement detector (LGMD) in the
locust's visual pathways, which represents high spike frequency to rapid
approaching objects. Building upon our previous models, in this paper we
propose a novel inhibition mechanism that is capable of adapting to different
levels of background complexity. This adaptive mechanism works effectively to
mediate the local inhibition strength and tune the temporal latency of local
excitation reaching the LGMD neuron. As a result, the proposed model is
effective to extract colliding cues from complex dynamic visual scenes. We
tested the proposed method using a range of stimuli including simulated
movements in grating backgrounds and shifting of a natural panoramic scene, as
well as vehicle crash video sequences. The experimental results demonstrate the
proposed method is feasible for fast collision perception in real-world
situations with potential applications in future autonomous vehicles.
| [
{
"created": "Wed, 3 Apr 2019 16:05:56 GMT",
"version": "v1"
}
] | 2019-04-04 | [
[
"Fu",
"Qinbing",
""
],
[
"Bellotto",
"Nicola",
""
],
[
"Wang",
"Huatian",
""
],
[
"Rind",
"F. Claire",
""
],
[
"Wang",
"Hongxin",
""
],
[
"Yue",
"Shigang",
""
]
] | This research addresses the challenging problem of visual collision detection in very complex and dynamic real physical scenes, specifically, the vehicle driving scenarios. This research takes inspiration from a large-field looming sensitive neuron, i.e., the lobula giant movement detector (LGMD) in the locust's visual pathways, which represents high spike frequency to rapid approaching objects. Building upon our previous models, in this paper we propose a novel inhibition mechanism that is capable of adapting to different levels of background complexity. This adaptive mechanism works effectively to mediate the local inhibition strength and tune the temporal latency of local excitation reaching the LGMD neuron. As a result, the proposed model is effective to extract colliding cues from complex dynamic visual scenes. We tested the proposed method using a range of stimuli including simulated movements in grating backgrounds and shifting of a natural panoramic scene, as well as vehicle crash video sequences. The experimental results demonstrate the proposed method is feasible for fast collision perception in real-world situations with potential applications in future autonomous vehicles. |
2012.07347 | Maxim Vashkevich | Maxim Vashkevich and Yulia Rushkevich | Classification of ALS patients based on acoustic analysis of sustained
vowel phonations | null | Biomedical Signal Processing and Control, Volume 65, March 2021,
102350 | 10.1016/j.bspc.2020.102350 | null | cs.SD cs.CL cs.LG eess.AS | http://creativecommons.org/licenses/by/4.0/ | Amyotrophic lateral sclerosis (ALS) is incurable neurological disorder with
rapidly progressive course. Common early symptoms of ALS are difficulty in
swallowing and speech. However, early acoustic manifestation of speech and
voice symptoms is very variable, that making their detection very challenging,
both by human specialists and automatic systems. This study presents an
approach to voice assessment for automatic system that separates healthy people
from patients with ALS. In particular, this work focus on analysing of sustain
phonation of vowels /a/ and /i/ to perform automatic classification of ALS
patients. A wide range of acoustic features such as MFCC, formants, jitter,
shimmer, vibrato, PPE, GNE, HNR, etc. were analysed. We also proposed a new set
of acoustic features for characterizing harmonic structure of the vowels.
Calculation of these features is based on pitch synchronized voice analysis. A
linear discriminant analysis (LDA) was used to classify the phonation produced
by patients with ALS and those by healthy individuals. Several algorithms of
feature selection were tested to find optimal feature subset for LDA model. The
study's experiments show that the most successful LDA model based on 32
features picked out by LASSO feature selection algorithm attains 99.7% accuracy
with 99.3% sensitivity and 99.9% specificity. Among the classifiers with a
small number of features, we can highlight LDA model with 5 features, which has
89.0% accuracy (87.5% sensitivity and 90.4% specificity).
| [
{
"created": "Mon, 14 Dec 2020 08:56:53 GMT",
"version": "v1"
},
{
"created": "Mon, 11 Jan 2021 08:41:07 GMT",
"version": "v2"
}
] | 2021-01-12 | [
[
"Vashkevich",
"Maxim",
""
],
[
"Rushkevich",
"Yulia",
""
]
] | Amyotrophic lateral sclerosis (ALS) is incurable neurological disorder with rapidly progressive course. Common early symptoms of ALS are difficulty in swallowing and speech. However, early acoustic manifestation of speech and voice symptoms is very variable, that making their detection very challenging, both by human specialists and automatic systems. This study presents an approach to voice assessment for automatic system that separates healthy people from patients with ALS. In particular, this work focus on analysing of sustain phonation of vowels /a/ and /i/ to perform automatic classification of ALS patients. A wide range of acoustic features such as MFCC, formants, jitter, shimmer, vibrato, PPE, GNE, HNR, etc. were analysed. We also proposed a new set of acoustic features for characterizing harmonic structure of the vowels. Calculation of these features is based on pitch synchronized voice analysis. A linear discriminant analysis (LDA) was used to classify the phonation produced by patients with ALS and those by healthy individuals. Several algorithms of feature selection were tested to find optimal feature subset for LDA model. The study's experiments show that the most successful LDA model based on 32 features picked out by LASSO feature selection algorithm attains 99.7% accuracy with 99.3% sensitivity and 99.9% specificity. Among the classifiers with a small number of features, we can highlight LDA model with 5 features, which has 89.0% accuracy (87.5% sensitivity and 90.4% specificity). |
2406.07547 | Xi Chen | Xi Chen, Yutong Feng, Mengting Chen, Yiyang Wang, Shilong Zhang, Yu
Liu, Yujun Shen, Hengshuang Zhao | Zero-shot Image Editing with Reference Imitation | https://xavierchen34.github.io/MimicBrush-Page | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | Image editing serves as a practical yet challenging task considering the
diverse demands from users, where one of the hardest parts is to precisely
describe how the edited image should look like. In this work, we present a new
form of editing, termed imitative editing, to help users exercise their
creativity more conveniently. Concretely, to edit an image region of interest,
users are free to directly draw inspiration from some in-the-wild references
(e.g., some relative pictures come across online), without having to cope with
the fit between the reference and the source. Such a design requires the system
to automatically figure out what to expect from the reference to perform the
editing. For this purpose, we propose a generative training framework, dubbed
MimicBrush, which randomly selects two frames from a video clip, masks some
regions of one frame, and learns to recover the masked regions using the
information from the other frame. That way, our model, developed from a
diffusion prior, is able to capture the semantic correspondence between
separate images in a self-supervised manner. We experimentally show the
effectiveness of our method under various test cases as well as its superiority
over existing alternatives. We also construct a benchmark to facilitate further
research.
| [
{
"created": "Tue, 11 Jun 2024 17:59:51 GMT",
"version": "v1"
}
] | 2024-06-12 | [
[
"Chen",
"Xi",
""
],
[
"Feng",
"Yutong",
""
],
[
"Chen",
"Mengting",
""
],
[
"Wang",
"Yiyang",
""
],
[
"Zhang",
"Shilong",
""
],
[
"Liu",
"Yu",
""
],
[
"Shen",
"Yujun",
""
],
[
"Zhao",
"Hengshuang",
""
]
] | Image editing serves as a practical yet challenging task considering the diverse demands from users, where one of the hardest parts is to precisely describe how the edited image should look like. In this work, we present a new form of editing, termed imitative editing, to help users exercise their creativity more conveniently. Concretely, to edit an image region of interest, users are free to directly draw inspiration from some in-the-wild references (e.g., some relative pictures come across online), without having to cope with the fit between the reference and the source. Such a design requires the system to automatically figure out what to expect from the reference to perform the editing. For this purpose, we propose a generative training framework, dubbed MimicBrush, which randomly selects two frames from a video clip, masks some regions of one frame, and learns to recover the masked regions using the information from the other frame. That way, our model, developed from a diffusion prior, is able to capture the semantic correspondence between separate images in a self-supervised manner. We experimentally show the effectiveness of our method under various test cases as well as its superiority over existing alternatives. We also construct a benchmark to facilitate further research. |
1904.03256 | Mohammad Sadegh Rasooli | Maryam Aminian, Mohammad Sadegh Rasooli, Mona Diab | Cross-Lingual Transfer of Semantic Roles: From Raw Text to Semantic
Roles | Accepted at the 13th International Conference on Computational
Semantics (IWCS 2019) | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe a transfer method based on annotation projection to develop a
dependency-based semantic role labeling system for languages for which no
supervised linguistic information other than parallel data is available. Unlike
previous work that presumes the availability of supervised features such as
lemmas, part-of-speech tags, and dependency parse trees, we only make use of
word and character features. Our deep model considers using character-based
representations as well as unsupervised stem embeddings to alleviate the need
for supervised features. Our experiments outperform a state-of-the-art method
that uses supervised lexico-syntactic features on 6 out of 7 languages in the
Universal Proposition Bank.
| [
{
"created": "Fri, 5 Apr 2019 20:04:04 GMT",
"version": "v1"
}
] | 2019-04-09 | [
[
"Aminian",
"Maryam",
""
],
[
"Rasooli",
"Mohammad Sadegh",
""
],
[
"Diab",
"Mona",
""
]
] | We describe a transfer method based on annotation projection to develop a dependency-based semantic role labeling system for languages for which no supervised linguistic information other than parallel data is available. Unlike previous work that presumes the availability of supervised features such as lemmas, part-of-speech tags, and dependency parse trees, we only make use of word and character features. Our deep model considers using character-based representations as well as unsupervised stem embeddings to alleviate the need for supervised features. Our experiments outperform a state-of-the-art method that uses supervised lexico-syntactic features on 6 out of 7 languages in the Universal Proposition Bank. |
2109.02914 | Junghyo Jo | Sungyeop Lee and Junghyo Jo | Scale-invariant representation of machine learning | null | null | 10.1103/PhysRevE.105.044306 | null | cs.LG cs.IT math.IT physics.data-an | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The success of machine learning has resulted from its structured
representation of data. Similar data have close internal representations as
compressed codes for classification or emerged labels for clustering. We
observe that the frequency of internal codes or labels follows power laws in
both supervised and unsupervised learning models. This scale-invariant
distribution implies that machine learning largely compresses frequent typical
data, and simultaneously, differentiates many atypical data as outliers. In
this study, we derive the process by which these power laws can naturally arise
in machine learning. In terms of information theory, the scale-invariant
representation corresponds to a maximally uncertain data grouping among
possible representations that guarantee a given learning accuracy.
| [
{
"created": "Tue, 7 Sep 2021 07:56:15 GMT",
"version": "v1"
},
{
"created": "Wed, 23 Mar 2022 08:11:08 GMT",
"version": "v2"
}
] | 2022-04-13 | [
[
"Lee",
"Sungyeop",
""
],
[
"Jo",
"Junghyo",
""
]
] | The success of machine learning has resulted from its structured representation of data. Similar data have close internal representations as compressed codes for classification or emerged labels for clustering. We observe that the frequency of internal codes or labels follows power laws in both supervised and unsupervised learning models. This scale-invariant distribution implies that machine learning largely compresses frequent typical data, and simultaneously, differentiates many atypical data as outliers. In this study, we derive the process by which these power laws can naturally arise in machine learning. In terms of information theory, the scale-invariant representation corresponds to a maximally uncertain data grouping among possible representations that guarantee a given learning accuracy. |
1404.0977 | Oren Weimann | Shay Mozes, Yahav Nussbaum, Oren Weimann | Faster Shortest Paths in Dense Distance Graphs, with Applications | null | null | null | null | cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We show how to combine two techniques for efficiently computing shortest
paths in directed planar graphs. The first is the linear-time shortest-path
algorithm of Henzinger, Klein, Subramanian, and Rao [STOC'94]. The second is
Fakcharoenphol and Rao's algorithm [FOCS'01] for emulating Dijkstra's algorithm
on the dense distance graph (DDG). A DDG is defined for a decomposition of a
planar graph $G$ into regions of at most $r$ vertices each, for some parameter
$r < n$. The vertex set of the DDG is the set of $\Theta(n/\sqrt r)$ vertices
of $G$ that belong to more than one region (boundary vertices). The DDG has
$\Theta(n)$ arcs, such that distances in the DDG are equal to the distances in
$G$. Fakcharoenphol and Rao's implementation of Dijkstra's algorithm on the DDG
(nicknamed FR-Dijkstra) runs in $O(n\log(n) r^{-1/2} \log r)$ time, and is a
key component in many state-of-the-art planar graph algorithms for shortest
paths, minimum cuts, and maximum flows. By combining these two techniques we
remove the $\log n$ dependency in the running time of the shortest-path
algorithm, making it $O(n r^{-1/2} \log^2r)$.
This work is part of a research agenda that aims to develop new techniques
that would lead to faster, possibly linear-time, algorithms for problems such
as minimum-cut, maximum-flow, and shortest paths with negative arc lengths. As
immediate applications, we show how to compute maximum flow in directed
weighted planar graphs in $O(n \log p)$ time, where $p$ is the minimum number
of edges on any path from the source to the sink. We also show how to compute
any part of the DDG that corresponds to a region with $r$ vertices and $k$
boundary vertices in $O(r \log k)$ time, which is faster than has been
previously known for small values of $k$.
| [
{
"created": "Thu, 3 Apr 2014 15:44:54 GMT",
"version": "v1"
}
] | 2014-04-04 | [
[
"Mozes",
"Shay",
""
],
[
"Nussbaum",
"Yahav",
""
],
[
"Weimann",
"Oren",
""
]
] | We show how to combine two techniques for efficiently computing shortest paths in directed planar graphs. The first is the linear-time shortest-path algorithm of Henzinger, Klein, Subramanian, and Rao [STOC'94]. The second is Fakcharoenphol and Rao's algorithm [FOCS'01] for emulating Dijkstra's algorithm on the dense distance graph (DDG). A DDG is defined for a decomposition of a planar graph $G$ into regions of at most $r$ vertices each, for some parameter $r < n$. The vertex set of the DDG is the set of $\Theta(n/\sqrt r)$ vertices of $G$ that belong to more than one region (boundary vertices). The DDG has $\Theta(n)$ arcs, such that distances in the DDG are equal to the distances in $G$. Fakcharoenphol and Rao's implementation of Dijkstra's algorithm on the DDG (nicknamed FR-Dijkstra) runs in $O(n\log(n) r^{-1/2} \log r)$ time, and is a key component in many state-of-the-art planar graph algorithms for shortest paths, minimum cuts, and maximum flows. By combining these two techniques we remove the $\log n$ dependency in the running time of the shortest-path algorithm, making it $O(n r^{-1/2} \log^2r)$. This work is part of a research agenda that aims to develop new techniques that would lead to faster, possibly linear-time, algorithms for problems such as minimum-cut, maximum-flow, and shortest paths with negative arc lengths. As immediate applications, we show how to compute maximum flow in directed weighted planar graphs in $O(n \log p)$ time, where $p$ is the minimum number of edges on any path from the source to the sink. We also show how to compute any part of the DDG that corresponds to a region with $r$ vertices and $k$ boundary vertices in $O(r \log k)$ time, which is faster than has been previously known for small values of $k$. |
1909.02564 | Jarom\'ir Janisch | Jarom\'ir Janisch, Tom\'a\v{s} Pevn\'y and Viliam Lis\'y | Classification with Costly Features as a Sequential Decision-Making
Problem | null | Machine Learning (2020): 1-29 | 10.1007/s10994-020-05874-8 | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This work focuses on a specific classification problem, where the information
about a sample is not readily available, but has to be acquired for a cost, and
there is a per-sample budget. Inspired by real-world use-cases, we analyze
average and hard variations of a directly specified budget. We postulate the
problem in its explicit formulation and then convert it into an equivalent MDP,
that can be solved with deep reinforcement learning. Also, we evaluate a
real-world inspired setting with sparse training dataset with missing features.
The presented method performs robustly well in all settings across several
distinct datasets, outperforming other prior-art algorithms. The method is
flexible, as showcased with all mentioned modifications and can be improved
with any domain independent advancement in RL.
| [
{
"created": "Thu, 5 Sep 2019 14:46:40 GMT",
"version": "v1"
}
] | 2020-03-05 | [
[
"Janisch",
"Jaromír",
""
],
[
"Pevný",
"Tomáš",
""
],
[
"Lisý",
"Viliam",
""
]
] | This work focuses on a specific classification problem, where the information about a sample is not readily available, but has to be acquired for a cost, and there is a per-sample budget. Inspired by real-world use-cases, we analyze average and hard variations of a directly specified budget. We postulate the problem in its explicit formulation and then convert it into an equivalent MDP, that can be solved with deep reinforcement learning. Also, we evaluate a real-world inspired setting with sparse training dataset with missing features. The presented method performs robustly well in all settings across several distinct datasets, outperforming other prior-art algorithms. The method is flexible, as showcased with all mentioned modifications and can be improved with any domain independent advancement in RL. |
1204.1420 | Abdullah Alshehab M. | Abdullah Alshehab, Chiu Tung Wu, Nao Kobayashi, Sikieng Sok, Shigeru
Shimamoto | Intra-bodyhybrid communication scheme for healthcare systems | International Journal on Bioinformatics & Biosciences (IJBB) Vol.2,
No.1, March 2012 | null | 10.5121/ijbb.2012.21011 | null | cs.ET | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Intra-body communication (IBC) is a type of Body Area Network (BAN)that
utilizes human body as the medium for data transmission. Thelow power
requirements of intra-body communication (IBC) as compared to near field
electromagnetic waves showed that it can be a suitable solution for Medical
Body Area Networks (MBANs) in a mobile health care system.In this paper, we
investigate the transmission characteristics of the human body as a conductor
of signals by considering different data transmission rates of multi-point to
point network in order to reduce overall power consumption of the
BAN.Furthermore, we utilize IBC and propose a new scheme to combines Slotted
ALOHA, TDMA, and Reservation ALOHA together to increase the throughput and
decrease the delay. By using our new hybrid scheme with the movable boundary
designed for health status monitoring, we are able to increase the efficiency
of data transmission by prioritizing the more critical data from the sensors.
| [
{
"created": "Fri, 6 Apr 2012 07:06:43 GMT",
"version": "v1"
}
] | 2012-04-09 | [
[
"Alshehab",
"Abdullah",
""
],
[
"Wu",
"Chiu Tung",
""
],
[
"Kobayashi",
"Nao",
""
],
[
"Sok",
"Sikieng",
""
],
[
"Shimamoto",
"Shigeru",
""
]
] | Intra-body communication (IBC) is a type of Body Area Network (BAN)that utilizes human body as the medium for data transmission. Thelow power requirements of intra-body communication (IBC) as compared to near field electromagnetic waves showed that it can be a suitable solution for Medical Body Area Networks (MBANs) in a mobile health care system.In this paper, we investigate the transmission characteristics of the human body as a conductor of signals by considering different data transmission rates of multi-point to point network in order to reduce overall power consumption of the BAN.Furthermore, we utilize IBC and propose a new scheme to combines Slotted ALOHA, TDMA, and Reservation ALOHA together to increase the throughput and decrease the delay. By using our new hybrid scheme with the movable boundary designed for health status monitoring, we are able to increase the efficiency of data transmission by prioritizing the more critical data from the sensors. |
2111.04264 | Chenglong Li | Chenglong Li, Tianhao Zhu, Lei Liu, Xiaonan Si, Zilin Fan, Sulan Zhai | Cross-Modal Object Tracking: Modality-Aware Representations and A
Unified Benchmark | In Submission | null | null | null | cs.CV | http://creativecommons.org/licenses/by/4.0/ | In many visual systems, visual tracking often bases on RGB image sequences,
in which some targets are invalid in low-light conditions, and tracking
performance is thus affected significantly. Introducing other modalities such
as depth and infrared data is an effective way to handle imaging limitations of
individual sources, but multi-modal imaging platforms usually require elaborate
designs and cannot be applied in many real-world applications at present.
Near-infrared (NIR) imaging becomes an essential part of many surveillance
cameras, whose imaging is switchable between RGB and NIR based on the light
intensity. These two modalities are heterogeneous with very different visual
properties and thus bring big challenges for visual tracking. However, existing
works have not studied this challenging problem. In this work, we address the
cross-modal object tracking problem and contribute a new video dataset,
including 654 cross-modal image sequences with over 481K frames in total, and
the average video length is more than 735 frames. To promote the research and
development of cross-modal object tracking, we propose a new algorithm, which
learns the modality-aware target representation to mitigate the appearance gap
between RGB and NIR modalities in the tracking process. It is plug-and-play and
could thus be flexibly embedded into different tracking frameworks. Extensive
experiments on the dataset are conducted, and we demonstrate the effectiveness
of the proposed algorithm in two representative tracking frameworks against 17
state-of-the-art tracking methods. We will release the dataset for free
academic usage, dataset download link and code will be released soon.
| [
{
"created": "Mon, 8 Nov 2021 03:58:55 GMT",
"version": "v1"
},
{
"created": "Thu, 11 Nov 2021 08:30:58 GMT",
"version": "v2"
}
] | 2021-11-12 | [
[
"Li",
"Chenglong",
""
],
[
"Zhu",
"Tianhao",
""
],
[
"Liu",
"Lei",
""
],
[
"Si",
"Xiaonan",
""
],
[
"Fan",
"Zilin",
""
],
[
"Zhai",
"Sulan",
""
]
] | In many visual systems, visual tracking often bases on RGB image sequences, in which some targets are invalid in low-light conditions, and tracking performance is thus affected significantly. Introducing other modalities such as depth and infrared data is an effective way to handle imaging limitations of individual sources, but multi-modal imaging platforms usually require elaborate designs and cannot be applied in many real-world applications at present. Near-infrared (NIR) imaging becomes an essential part of many surveillance cameras, whose imaging is switchable between RGB and NIR based on the light intensity. These two modalities are heterogeneous with very different visual properties and thus bring big challenges for visual tracking. However, existing works have not studied this challenging problem. In this work, we address the cross-modal object tracking problem and contribute a new video dataset, including 654 cross-modal image sequences with over 481K frames in total, and the average video length is more than 735 frames. To promote the research and development of cross-modal object tracking, we propose a new algorithm, which learns the modality-aware target representation to mitigate the appearance gap between RGB and NIR modalities in the tracking process. It is plug-and-play and could thus be flexibly embedded into different tracking frameworks. Extensive experiments on the dataset are conducted, and we demonstrate the effectiveness of the proposed algorithm in two representative tracking frameworks against 17 state-of-the-art tracking methods. We will release the dataset for free academic usage, dataset download link and code will be released soon. |
1907.10322 | Simon Tamayo Giraldo | Sarah Manard (CAOR), Nicolas Vergos (CAOR), Simon Tamayo (CAOR),
Fr\'ed\'eric Fontane (CAOR) | Electronic health record in the era of industry 4.0: the French example | null | International Conference e-Health 2019, Jul 2019, Porto, Portugal | null | null | cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The recent implementation of the Electronic Health Record (EHR) in France is
part of a more general process of digitizing information flows, as the world
enters the fourth industrial revolution in a phenomenon known as Industry 4.0.
Behind this concept lies the concern to allow Man to remain permanently in
control of his destiny, despite an increasingly interconnected world (Internet
of Things, cooperative robots, augmented reality, etc.). Accordingly, the
implementation of EHR must guarantee the respect for the private life of each
citizen. From this perspective, healthcare professionals will therefore have to
constantly ensure the protection of medical confidentiality during Electronic
Data Interchange (EDI). This paper summarises the current state of the use of
EHR in France. Based on a survey conducted by the European Commission to assess
the deployment of digitalisation in the health sector in EU countries, this
article aims to highlight the opportunities and perspectives that Industry 4.0
could bring to the health sector in France. However, this study also identifies
a number of limits related to the application of such a system, the first of
which is cyber threat or transhumanism. To this end, a SWOT matrix identifies
the strengths and weaknesses related to the implementation of the French EHR.
| [
{
"created": "Wed, 24 Jul 2019 09:24:24 GMT",
"version": "v1"
},
{
"created": "Thu, 25 Jul 2019 09:00:12 GMT",
"version": "v2"
}
] | 2019-07-26 | [
[
"Manard",
"Sarah",
"",
"CAOR"
],
[
"Vergos",
"Nicolas",
"",
"CAOR"
],
[
"Tamayo",
"Simon",
"",
"CAOR"
],
[
"Fontane",
"Frédéric",
"",
"CAOR"
]
] | The recent implementation of the Electronic Health Record (EHR) in France is part of a more general process of digitizing information flows, as the world enters the fourth industrial revolution in a phenomenon known as Industry 4.0. Behind this concept lies the concern to allow Man to remain permanently in control of his destiny, despite an increasingly interconnected world (Internet of Things, cooperative robots, augmented reality, etc.). Accordingly, the implementation of EHR must guarantee the respect for the private life of each citizen. From this perspective, healthcare professionals will therefore have to constantly ensure the protection of medical confidentiality during Electronic Data Interchange (EDI). This paper summarises the current state of the use of EHR in France. Based on a survey conducted by the European Commission to assess the deployment of digitalisation in the health sector in EU countries, this article aims to highlight the opportunities and perspectives that Industry 4.0 could bring to the health sector in France. However, this study also identifies a number of limits related to the application of such a system, the first of which is cyber threat or transhumanism. To this end, a SWOT matrix identifies the strengths and weaknesses related to the implementation of the French EHR. |
2109.01879 | Anindya Mondal | Anindya Mondal, Mayukhmali Das | Moving Object Detection for Event-based Vision using k-means Clustering | Nine pages, five figures, Published in 2021 IEEE 8th Uttar Pradesh
Section International Conference on Electrical, Electronics and Computer
Engineering (UPCON) | null | 10.1109/UPCON52273.2021.9667636 | null | cs.CV cs.AI eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Moving object detection is important in computer vision. Event-based cameras
are bio-inspired cameras that work by mimicking the working of the human eye.
These cameras have multiple advantages over conventional frame-based cameras,
like reduced latency, HDR, reduced motion blur during high motion, low power
consumption, etc. In spite of these advantages, event-based cameras are
noise-sensitive and have low resolution. Moreover, the task of moving object
detection in these cameras is difficult, as event-based sensors lack useful
visual features like texture and color. In this paper, we investigate the
application of the k-means clustering technique in detecting moving objects in
event-based data.
| [
{
"created": "Sat, 4 Sep 2021 14:43:14 GMT",
"version": "v1"
},
{
"created": "Fri, 1 Oct 2021 16:06:17 GMT",
"version": "v2"
},
{
"created": "Mon, 8 Nov 2021 08:24:19 GMT",
"version": "v3"
},
{
"created": "Tue, 11 Jan 2022 21:03:51 GMT",
"version": "v4"
}
] | 2022-01-13 | [
[
"Mondal",
"Anindya",
""
],
[
"Das",
"Mayukhmali",
""
]
] | Moving object detection is important in computer vision. Event-based cameras are bio-inspired cameras that work by mimicking the working of the human eye. These cameras have multiple advantages over conventional frame-based cameras, like reduced latency, HDR, reduced motion blur during high motion, low power consumption, etc. In spite of these advantages, event-based cameras are noise-sensitive and have low resolution. Moreover, the task of moving object detection in these cameras is difficult, as event-based sensors lack useful visual features like texture and color. In this paper, we investigate the application of the k-means clustering technique in detecting moving objects in event-based data. |
0806.4553 | Viorica Sofronie-Stokkermans | Viorica Sofronie-Stokkermans | Interpolation in local theory extensions | 31 pages, 1 figure | Logical Methods in Computer Science, Volume 4, Issue 4 (October
17, 2008) lmcs:1143 | 10.2168/LMCS-4(4:1)2008 | null | cs.LO cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we study interpolation in local extensions of a base theory. We
identify situations in which it is possible to obtain interpolants in a
hierarchical manner, by using a prover and a procedure for generating
interpolants in the base theory as black-boxes. We present several examples of
theory extensions in which interpolants can be computed this way, and discuss
applications in verification, knowledge representation, and modular reasoning
in combinations of local theories.
| [
{
"created": "Fri, 27 Jun 2008 15:51:02 GMT",
"version": "v1"
},
{
"created": "Thu, 16 Oct 2008 22:01:02 GMT",
"version": "v2"
}
] | 2015-07-01 | [
[
"Sofronie-Stokkermans",
"Viorica",
""
]
] | In this paper we study interpolation in local extensions of a base theory. We identify situations in which it is possible to obtain interpolants in a hierarchical manner, by using a prover and a procedure for generating interpolants in the base theory as black-boxes. We present several examples of theory extensions in which interpolants can be computed this way, and discuss applications in verification, knowledge representation, and modular reasoning in combinations of local theories. |
2303.07740 | Yang Bai | Min Cao, Yang Bai, Jingyao Wang, Ziqiang Cao, Liqiang Nie, Min Zhang | Efficient Image-Text Retrieval via Keyword-Guided Pre-Screening | 11 pages, 7 figures, 6 tables | null | null | null | cs.CV cs.CL | http://creativecommons.org/licenses/by/4.0/ | Under the flourishing development in performance, current image-text
retrieval methods suffer from $N$-related time complexity, which hinders their
application in practice. Targeting at efficiency improvement, this paper
presents a simple and effective keyword-guided pre-screening framework for the
image-text retrieval. Specifically, we convert the image and text data into the
keywords and perform the keyword matching across modalities to exclude a large
number of irrelevant gallery samples prior to the retrieval network. For the
keyword prediction, we transfer it into a multi-label classification problem
and propose a multi-task learning scheme by appending the multi-label
classifiers to the image-text retrieval network to achieve a lightweight and
high-performance keyword prediction. For the keyword matching, we introduce the
inverted index in the search engine and create a win-win situation on both time
and space complexities for the pre-screening. Extensive experiments on two
widely-used datasets, i.e., Flickr30K and MS-COCO, verify the effectiveness of
the proposed framework. The proposed framework equipped with only two embedding
layers achieves $O(1)$ querying time complexity, while improving the retrieval
efficiency and keeping its performance, when applied prior to the common
image-text retrieval methods. Our code will be released.
| [
{
"created": "Tue, 14 Mar 2023 09:36:42 GMT",
"version": "v1"
}
] | 2023-03-15 | [
[
"Cao",
"Min",
""
],
[
"Bai",
"Yang",
""
],
[
"Wang",
"Jingyao",
""
],
[
"Cao",
"Ziqiang",
""
],
[
"Nie",
"Liqiang",
""
],
[
"Zhang",
"Min",
""
]
] | Under the flourishing development in performance, current image-text retrieval methods suffer from $N$-related time complexity, which hinders their application in practice. Targeting at efficiency improvement, this paper presents a simple and effective keyword-guided pre-screening framework for the image-text retrieval. Specifically, we convert the image and text data into the keywords and perform the keyword matching across modalities to exclude a large number of irrelevant gallery samples prior to the retrieval network. For the keyword prediction, we transfer it into a multi-label classification problem and propose a multi-task learning scheme by appending the multi-label classifiers to the image-text retrieval network to achieve a lightweight and high-performance keyword prediction. For the keyword matching, we introduce the inverted index in the search engine and create a win-win situation on both time and space complexities for the pre-screening. Extensive experiments on two widely-used datasets, i.e., Flickr30K and MS-COCO, verify the effectiveness of the proposed framework. The proposed framework equipped with only two embedding layers achieves $O(1)$ querying time complexity, while improving the retrieval efficiency and keeping its performance, when applied prior to the common image-text retrieval methods. Our code will be released. |
0911.3343 | Nicolai Kuntze | Nicolai Kuntze, Juergen Repp, Hervais Simo Fhom, Andreas Fuchs,
Ine-Saf Benaissa | Final Architecture Specification of security, privacy, and incentive
mechanisms | Delieverable of the EU FP7 project Nanodatacenters | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this document, we define the NADA security architecture based on refined
use case scenarios, a derived high level model and security analysis. For the
architecure design and verification we are applying the well known STRIDE
model.
| [
{
"created": "Tue, 17 Nov 2009 15:58:10 GMT",
"version": "v1"
}
] | 2009-11-18 | [
[
"Kuntze",
"Nicolai",
""
],
[
"Repp",
"Juergen",
""
],
[
"Fhom",
"Hervais Simo",
""
],
[
"Fuchs",
"Andreas",
""
],
[
"Benaissa",
"Ine-Saf",
""
]
] | In this document, we define the NADA security architecture based on refined use case scenarios, a derived high level model and security analysis. For the architecure design and verification we are applying the well known STRIDE model. |
1807.01545 | Christian H\"ager | Christian H\"ager, Henry D. Pfister | Wideband Time-Domain Digital Backpropagation via Subband Processing and
Deep Learning | 3 pages, 3 figurs | null | null | null | cs.IT math.IT stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a low-complexity sub-banded DSP architecture for digital
backpropagation where the walk-off effect is compensated using simple delay
elements. For a simulated 96-Gbaud signal and 2500 km optical link, our method
achieves a 2.8 dB SNR improvement over linear equalization.
| [
{
"created": "Wed, 4 Jul 2018 12:39:25 GMT",
"version": "v1"
}
] | 2018-07-05 | [
[
"Häger",
"Christian",
""
],
[
"Pfister",
"Henry D.",
""
]
] | We propose a low-complexity sub-banded DSP architecture for digital backpropagation where the walk-off effect is compensated using simple delay elements. For a simulated 96-Gbaud signal and 2500 km optical link, our method achieves a 2.8 dB SNR improvement over linear equalization. |
2202.05347 | T\'ulio Marcondes Moreira | T\'ulio Marcondes Moreira, Jackson Geraldo de Faria Jr, Pedro O.S.
Vaz-de-Melo and Gilberto Medeiros-Ribeiro | Development and Validation of an AI-Driven Model for the La Rance Tidal
Barrage: A Generalisable Case Study | 30 pages, 22 figures and 6 tables | null | null | null | cs.LG | http://creativecommons.org/licenses/by-nc-nd/4.0/ | In this work, an AI-Driven (autonomous) model representation of the La Rance
tidal barrage was developed using novel parametrisation and Deep Reinforcement
Learning (DRL) techniques. Our model results were validated with experimental
measurements, yielding the first Tidal Range Structure (TRS) model validated
against a constructed tidal barrage and made available to academics. In order
to proper model La Rance, parametrisation methodologies were developed for
simulating (i) turbines (in pumping and power generation modes), (ii)
transition ramp functions (for opening and closing hydraulic structures) and
(iii) equivalent lagoon wetted area. Furthermore, an updated DRL method was
implemented for optimising the operation of the hydraulic structures that
compose La Rance. The achieved objective of this work was to verify the
capabilities of an AI-Driven TRS model to appropriately predict (i) turbine
power and (ii) lagoon water level variations. In addition, the observed
operational strategy and yearly energy output of our AI-Driven model appeared
to be comparable with those reported for the La Rance tidal barrage. The
outcomes of this work (developed methodologies and DRL implementations) are
generalisable and can be applied to other TRS projects. Furthermore, this work
provided insights which allow for more realistic simulation of TRS operation,
enabled through our AI-Driven model.
| [
{
"created": "Thu, 10 Feb 2022 22:02:52 GMT",
"version": "v1"
}
] | 2022-02-14 | [
[
"Moreira",
"Túlio Marcondes",
""
],
[
"Faria",
"Jackson Geraldo de",
"Jr"
],
[
"Vaz-de-Melo",
"Pedro O. S.",
""
],
[
"Medeiros-Ribeiro",
"Gilberto",
""
]
] | In this work, an AI-Driven (autonomous) model representation of the La Rance tidal barrage was developed using novel parametrisation and Deep Reinforcement Learning (DRL) techniques. Our model results were validated with experimental measurements, yielding the first Tidal Range Structure (TRS) model validated against a constructed tidal barrage and made available to academics. In order to proper model La Rance, parametrisation methodologies were developed for simulating (i) turbines (in pumping and power generation modes), (ii) transition ramp functions (for opening and closing hydraulic structures) and (iii) equivalent lagoon wetted area. Furthermore, an updated DRL method was implemented for optimising the operation of the hydraulic structures that compose La Rance. The achieved objective of this work was to verify the capabilities of an AI-Driven TRS model to appropriately predict (i) turbine power and (ii) lagoon water level variations. In addition, the observed operational strategy and yearly energy output of our AI-Driven model appeared to be comparable with those reported for the La Rance tidal barrage. The outcomes of this work (developed methodologies and DRL implementations) are generalisable and can be applied to other TRS projects. Furthermore, this work provided insights which allow for more realistic simulation of TRS operation, enabled through our AI-Driven model. |
2207.05267 | Bo Wang | Haiqing Hao (1), Zhongwang Pang (1 and 2), Guan Wang (1 and 2) and Bo
Wang (1 and 2) ((1) State Key Laboratory of Precision Measurement Technology
and Instruments, Department of Precision Instrument, Tsinghua University,
Beijing, China, (2) Key Laboratory of Photonic Control Technology (Tsinghua
University), Ministry of Education, Beijing, China) | Indoor optical fiber eavesdropping approach and its avoidance | 8 pages, 4 figures, submitted to Optics Express | null | 10.1364/OE.470529 | null | cs.SD eess.AS physics.ins-det physics.optics | http://creativecommons.org/licenses/by-nc-nd/4.0/ | The optical fiber network has become a worldwide infrastructure. In addition
to the basic functions in telecommunication, its sensing ability has attracted
more and more attention. In this paper, we discuss the risk of household fiber
being used for eavesdropping and demonstrate its performance in the lab. Using
a 3-meter tail fiber in front of the household optical modem, voices of normal
human speech can be eavesdropped by a laser interferometer and recovered 1.1 km
away. The detection distance limit and system noise are analyzed
quantitatively. We also give some practical ways to prevent eavesdropping
through household fiber.
| [
{
"created": "Tue, 12 Jul 2022 02:31:34 GMT",
"version": "v1"
},
{
"created": "Wed, 3 Aug 2022 13:58:31 GMT",
"version": "v2"
}
] | 2022-10-05 | [
[
"Hao",
"Haiqing",
"",
"1 and 2"
],
[
"Pang",
"Zhongwang",
"",
"1 and 2"
],
[
"Wang",
"Guan",
"",
"1 and 2"
],
[
"Wang",
"Bo",
"",
"1 and 2"
]
] | The optical fiber network has become a worldwide infrastructure. In addition to the basic functions in telecommunication, its sensing ability has attracted more and more attention. In this paper, we discuss the risk of household fiber being used for eavesdropping and demonstrate its performance in the lab. Using a 3-meter tail fiber in front of the household optical modem, voices of normal human speech can be eavesdropped by a laser interferometer and recovered 1.1 km away. The detection distance limit and system noise are analyzed quantitatively. We also give some practical ways to prevent eavesdropping through household fiber. |
2007.11849 | Chen-Yu Wei | Chen-Yu Wei, Mehdi Jafarnia-Jahromi, Haipeng Luo, Rahul Jain | Learning Infinite-horizon Average-reward MDPs with Linear Function
Approximation | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We develop several new algorithms for learning Markov Decision Processes in
an infinite-horizon average-reward setting with linear function approximation.
Using the optimism principle and assuming that the MDP has a linear structure,
we first propose a computationally inefficient algorithm with optimal
$\widetilde{O}(\sqrt{T})$ regret and another computationally efficient variant
with $\widetilde{O}(T^{3/4})$ regret, where $T$ is the number of interactions.
Next, taking inspiration from adversarial linear bandits, we develop yet
another efficient algorithm with $\widetilde{O}(\sqrt{T})$ regret under a
different set of assumptions, improving the best existing result by Hao et al.
(2020) with $\widetilde{O}(T^{2/3})$ regret. Moreover, we draw a connection
between this algorithm and the Natural Policy Gradient algorithm proposed by
Kakade (2002), and show that our analysis improves the sample complexity bound
recently given by Agarwal et al. (2020).
| [
{
"created": "Thu, 23 Jul 2020 08:23:44 GMT",
"version": "v1"
},
{
"created": "Mon, 26 Apr 2021 09:12:03 GMT",
"version": "v2"
}
] | 2021-04-27 | [
[
"Wei",
"Chen-Yu",
""
],
[
"Jafarnia-Jahromi",
"Mehdi",
""
],
[
"Luo",
"Haipeng",
""
],
[
"Jain",
"Rahul",
""
]
] | We develop several new algorithms for learning Markov Decision Processes in an infinite-horizon average-reward setting with linear function approximation. Using the optimism principle and assuming that the MDP has a linear structure, we first propose a computationally inefficient algorithm with optimal $\widetilde{O}(\sqrt{T})$ regret and another computationally efficient variant with $\widetilde{O}(T^{3/4})$ regret, where $T$ is the number of interactions. Next, taking inspiration from adversarial linear bandits, we develop yet another efficient algorithm with $\widetilde{O}(\sqrt{T})$ regret under a different set of assumptions, improving the best existing result by Hao et al. (2020) with $\widetilde{O}(T^{2/3})$ regret. Moreover, we draw a connection between this algorithm and the Natural Policy Gradient algorithm proposed by Kakade (2002), and show that our analysis improves the sample complexity bound recently given by Agarwal et al. (2020). |
2010.09409 | Juan Jos\'e G\'omez Rodr\'iguez | Juan J. G\'omez Rodr\'iguez, Jos\'e Lamarca, Javier Morlana, Juan D.
Tard\'os, Jos\'e M. M. Montiel | SD-DefSLAM: Semi-Direct Monocular SLAM for Deformable and Intracorporeal
Scenes | 10 pages, 8 figures. Submitted to RA-L with option to ICRA 2021.
Associated video: https://youtu.be/gkcC0IR3X6A | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Conventional SLAM techniques strongly rely on scene rigidity to solve data
association, ignoring dynamic parts of the scene. In this work we present
Semi-Direct DefSLAM (SD-DefSLAM), a novel monocular deformable SLAM method able
to map highly deforming environments, built on top of DefSLAM. To robustly
solve data association in challenging deforming scenes, SD-DefSLAM combines
direct and indirect methods: an enhanced illumination-invariant Lucas-Kanade
tracker for data association, geometric Bundle Adjustment for pose and
deformable map estimation, and bag-of-words based on feature descriptors for
camera relocation. Dynamic objects are detected and segmented-out using a CNN
trained for the specific application domain. We thoroughly evaluate our system
in two public datasets. The mandala dataset is a SLAM benchmark with
increasingly aggressive deformations. The Hamlyn dataset contains
intracorporeal sequences that pose serious real-life challenges beyond
deformation like weak texture, specular reflections, surgical tools and
occlusions. Our results show that SD-DefSLAM outperforms DefSLAM in point
tracking, reconstruction accuracy and scale drift thanks to the improvement in
all the data association steps, being the first system able to robustly perform
SLAM inside the human body.
| [
{
"created": "Mon, 19 Oct 2020 12:07:07 GMT",
"version": "v1"
}
] | 2020-10-20 | [
[
"Rodríguez",
"Juan J. Gómez",
""
],
[
"Lamarca",
"José",
""
],
[
"Morlana",
"Javier",
""
],
[
"Tardós",
"Juan D.",
""
],
[
"Montiel",
"José M. M.",
""
]
] | Conventional SLAM techniques strongly rely on scene rigidity to solve data association, ignoring dynamic parts of the scene. In this work we present Semi-Direct DefSLAM (SD-DefSLAM), a novel monocular deformable SLAM method able to map highly deforming environments, built on top of DefSLAM. To robustly solve data association in challenging deforming scenes, SD-DefSLAM combines direct and indirect methods: an enhanced illumination-invariant Lucas-Kanade tracker for data association, geometric Bundle Adjustment for pose and deformable map estimation, and bag-of-words based on feature descriptors for camera relocation. Dynamic objects are detected and segmented-out using a CNN trained for the specific application domain. We thoroughly evaluate our system in two public datasets. The mandala dataset is a SLAM benchmark with increasingly aggressive deformations. The Hamlyn dataset contains intracorporeal sequences that pose serious real-life challenges beyond deformation like weak texture, specular reflections, surgical tools and occlusions. Our results show that SD-DefSLAM outperforms DefSLAM in point tracking, reconstruction accuracy and scale drift thanks to the improvement in all the data association steps, being the first system able to robustly perform SLAM inside the human body. |
2403.06189 | Dai Yuqin | Yuqin Dai, Wanlu Zhu, Ronghui Li, Zeping Ren, Xiangzheng Zhou, Xiu Li,
Jun Li, Jian Yang | Harmonious Group Choreography with Trajectory-Controllable Diffusion | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Creating group choreography from music has gained attention in cultural
entertainment and virtual reality, aiming to coordinate visually cohesive and
diverse group movements. Despite increasing interest, recent works face
challenges in achieving aesthetically appealing choreography, primarily for two
key issues: multi-dancer collision and single-dancer foot slide. To address
these issues, we propose a Trajectory-Controllable Diffusion (TCDiff), a novel
approach that harnesses non-overlapping trajectories to facilitate coherent
dance movements. Specifically, to tackle dancer collisions, we introduce a
Dance-Beat Navigator capable of generating trajectories for multiple dancers
based on the music, complemented by a Distance-Consistency loss to maintain
appropriate spacing among trajectories within a reasonable threshold. To
mitigate foot sliding, we present a Footwork Adaptor that utilizes trajectory
displacement from adjacent frames to enable flexible footwork, coupled with a
Relative Forward-Kinematic loss to adjust the positioning of individual
dancers' root nodes and joints. Extensive experiments demonstrate that our
method achieves state-of-the-art results.
| [
{
"created": "Sun, 10 Mar 2024 12:11:34 GMT",
"version": "v1"
},
{
"created": "Thu, 6 Jun 2024 08:19:12 GMT",
"version": "v2"
},
{
"created": "Wed, 14 Aug 2024 02:38:55 GMT",
"version": "v3"
}
] | 2024-08-15 | [
[
"Dai",
"Yuqin",
""
],
[
"Zhu",
"Wanlu",
""
],
[
"Li",
"Ronghui",
""
],
[
"Ren",
"Zeping",
""
],
[
"Zhou",
"Xiangzheng",
""
],
[
"Li",
"Xiu",
""
],
[
"Li",
"Jun",
""
],
[
"Yang",
"Jian",
""
]
] | Creating group choreography from music has gained attention in cultural entertainment and virtual reality, aiming to coordinate visually cohesive and diverse group movements. Despite increasing interest, recent works face challenges in achieving aesthetically appealing choreography, primarily for two key issues: multi-dancer collision and single-dancer foot slide. To address these issues, we propose a Trajectory-Controllable Diffusion (TCDiff), a novel approach that harnesses non-overlapping trajectories to facilitate coherent dance movements. Specifically, to tackle dancer collisions, we introduce a Dance-Beat Navigator capable of generating trajectories for multiple dancers based on the music, complemented by a Distance-Consistency loss to maintain appropriate spacing among trajectories within a reasonable threshold. To mitigate foot sliding, we present a Footwork Adaptor that utilizes trajectory displacement from adjacent frames to enable flexible footwork, coupled with a Relative Forward-Kinematic loss to adjust the positioning of individual dancers' root nodes and joints. Extensive experiments demonstrate that our method achieves state-of-the-art results. |
2304.02572 | Zheqing Zhu | Hongbo Guo, Ruben Naeff, Alex Nikulkov, Zheqing Zhu | Evaluating Online Bandit Exploration In Large-Scale Recommender System | null | null | null | null | cs.IR cs.AI cs.LG cs.SI | http://creativecommons.org/licenses/by/4.0/ | Bandit learning has been an increasingly popular design choice for
recommender system. Despite the strong interest in bandit learning from the
community, there remains multiple bottlenecks that prevent many bandit learning
approaches from productionalization. One major bottleneck is how to test the
effectiveness of bandit algorithm with fairness and without data leakage.
Different from supervised learning algorithms, bandit learning algorithms
emphasize greatly on the data collection process through their explorative
nature. Such explorative behavior may induce unfair evaluation in a classic A/B
test setting. In this work, we apply upper confidence bound (UCB) to our large
scale short video recommender system and present a test framework for the
production bandit learning life-cycle with a new set of metrics. Extensive
experiment results show that our experiment design is able to fairly evaluate
the performance of bandit learning in the recommender system.
| [
{
"created": "Wed, 5 Apr 2023 16:44:36 GMT",
"version": "v1"
},
{
"created": "Thu, 22 Jun 2023 03:41:43 GMT",
"version": "v2"
},
{
"created": "Sun, 30 Jul 2023 08:29:55 GMT",
"version": "v3"
}
] | 2023-08-01 | [
[
"Guo",
"Hongbo",
""
],
[
"Naeff",
"Ruben",
""
],
[
"Nikulkov",
"Alex",
""
],
[
"Zhu",
"Zheqing",
""
]
] | Bandit learning has been an increasingly popular design choice for recommender system. Despite the strong interest in bandit learning from the community, there remains multiple bottlenecks that prevent many bandit learning approaches from productionalization. One major bottleneck is how to test the effectiveness of bandit algorithm with fairness and without data leakage. Different from supervised learning algorithms, bandit learning algorithms emphasize greatly on the data collection process through their explorative nature. Such explorative behavior may induce unfair evaluation in a classic A/B test setting. In this work, we apply upper confidence bound (UCB) to our large scale short video recommender system and present a test framework for the production bandit learning life-cycle with a new set of metrics. Extensive experiment results show that our experiment design is able to fairly evaluate the performance of bandit learning in the recommender system. |
1308.0037 | Ryan Williams | Ryan K. Williams, Andrea Gasparri, and Bhaskar Krishnamachari | Route Swarm: Wireless Network Optimization through Mobility | 9 pages, 4 figures, submitted to the IEEE International Conference on
Intelligent Robots and Systems (IROS) 2014 | null | 10.1109/IROS.2014.6943092 | null | cs.SY cs.MA cs.NI cs.RO math.OC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we demonstrate a novel hybrid architecture for coordinating
networked robots in sensing and information routing applications. The proposed
INformation and Sensing driven PhysIcally REconfigurable robotic network
(INSPIRE), consists of a Physical Control Plane (PCP) which commands agent
position, and an Information Control Plane (ICP) which regulates information
flow towards communication/sensing objectives. We describe an instantiation
where a mobile robotic network is dynamically reconfigured to ensure high
quality routes between static wireless nodes, which act as source/destination
pairs for information flow. The ICP commands the robots towards evenly
distributed inter-flow allocations, with intra-flow configurations that
maximize route quality. The PCP then guides the robots via potential-based
control to reconfigure according to ICP commands. This formulation, deemed
Route Swarm, decouples information flow and physical control, generating a
feedback between routing and sensing needs and robotic configuration. We
demonstrate our propositions through simulation under a realistic wireless
network regime.
| [
{
"created": "Wed, 31 Jul 2013 20:47:14 GMT",
"version": "v1"
},
{
"created": "Tue, 3 Sep 2013 21:16:47 GMT",
"version": "v2"
},
{
"created": "Fri, 7 Feb 2014 02:24:13 GMT",
"version": "v3"
}
] | 2016-11-18 | [
[
"Williams",
"Ryan K.",
""
],
[
"Gasparri",
"Andrea",
""
],
[
"Krishnamachari",
"Bhaskar",
""
]
] | In this paper, we demonstrate a novel hybrid architecture for coordinating networked robots in sensing and information routing applications. The proposed INformation and Sensing driven PhysIcally REconfigurable robotic network (INSPIRE), consists of a Physical Control Plane (PCP) which commands agent position, and an Information Control Plane (ICP) which regulates information flow towards communication/sensing objectives. We describe an instantiation where a mobile robotic network is dynamically reconfigured to ensure high quality routes between static wireless nodes, which act as source/destination pairs for information flow. The ICP commands the robots towards evenly distributed inter-flow allocations, with intra-flow configurations that maximize route quality. The PCP then guides the robots via potential-based control to reconfigure according to ICP commands. This formulation, deemed Route Swarm, decouples information flow and physical control, generating a feedback between routing and sensing needs and robotic configuration. We demonstrate our propositions through simulation under a realistic wireless network regime. |
1705.08362 | Thorsten Wi{\ss}mann | Ulrich Dorsch, Stefan Milius, Lutz Schr\"oder, Thorsten Wi{\ss}mann | Efficient Coalgebraic Partition Refinement | null | null | null | null | cs.DS cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a generic partition refinement algorithm that quotients
coalgebraic systems by behavioural equivalence, an important task in reactive
verification; coalgebraic generality implies in particular that we cover not
only classical relational systems but also various forms of weighted systems.
Under assumptions on the type functor that allow representing its finite
coalgebras in terms of nodes and edges, our algorithm runs in time
$\mathcal{O}(m\cdot \log n)$ where $n$ and $m$ are the numbers of nodes and
edges, respectively. Instances of our generic algorithm thus match the runtime
of the best known algorithms for unlabelled transition systems, Markov chains,
and deterministic automata (with fixed alphabets), and improve the best known
algorithms for Segala systems.
| [
{
"created": "Tue, 23 May 2017 15:31:59 GMT",
"version": "v1"
},
{
"created": "Sat, 8 Jul 2017 09:53:47 GMT",
"version": "v2"
},
{
"created": "Thu, 13 Jul 2017 10:49:21 GMT",
"version": "v3"
},
{
"created": "Mon, 9 Oct 2017 10:19:12 GMT",
"version": "v4"
}
] | 2017-10-10 | [
[
"Dorsch",
"Ulrich",
""
],
[
"Milius",
"Stefan",
""
],
[
"Schröder",
"Lutz",
""
],
[
"Wißmann",
"Thorsten",
""
]
] | We present a generic partition refinement algorithm that quotients coalgebraic systems by behavioural equivalence, an important task in reactive verification; coalgebraic generality implies in particular that we cover not only classical relational systems but also various forms of weighted systems. Under assumptions on the type functor that allow representing its finite coalgebras in terms of nodes and edges, our algorithm runs in time $\mathcal{O}(m\cdot \log n)$ where $n$ and $m$ are the numbers of nodes and edges, respectively. Instances of our generic algorithm thus match the runtime of the best known algorithms for unlabelled transition systems, Markov chains, and deterministic automata (with fixed alphabets), and improve the best known algorithms for Segala systems. |
1204.4909 | M. Rizwan Jameel Qureshi Dr. | M. Rizwan Jameel Qureshi and Waseem Qureshi | Evaluation of the Design Metric to Reduce the Number of Defects in
Software Development | 9 Pages | International Journal of Information Technology and Computer
Science (IJITCS), Vol. 4/4, pp. 9-17, April 2012 | 10.5815/ijitcs.2012.04.02 | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Software design is one of the most important and key activities in the system
development life cycle (SDLC) phase that ensures the quality of software.
Different key areas of design are very vital to be taken into consideration
while designing software. Software design describes how the software system is
decomposed and managed in smaller components. Object-oriented (OO) paradigm has
facilitated software industry with more reliable and manageable software and
its design. The quality of the software design can be measured through
different metrics such as Chidamber and Kemerer (CK) design metrics, Mood
Metrics & Lorenz and Kidd metrics. CK metrics is one of the oldest and most
reliable metrics among all metrics available to software industry to evaluate
OO design. This paper presents an evaluation of CK metrics to propose an
improved CK design metrics values to reduce the defects during software design
phase in software. This paper will also describe that whether a significant
effect of any CK design metrics exists on total number of defects per module or
not. This is achieved by conducting survey in two software development
companies.
| [
{
"created": "Sun, 22 Apr 2012 16:35:41 GMT",
"version": "v1"
}
] | 2012-04-24 | [
[
"Qureshi",
"M. Rizwan Jameel",
""
],
[
"Qureshi",
"Waseem",
""
]
] | Software design is one of the most important and key activities in the system development life cycle (SDLC) phase that ensures the quality of software. Different key areas of design are very vital to be taken into consideration while designing software. Software design describes how the software system is decomposed and managed in smaller components. Object-oriented (OO) paradigm has facilitated software industry with more reliable and manageable software and its design. The quality of the software design can be measured through different metrics such as Chidamber and Kemerer (CK) design metrics, Mood Metrics & Lorenz and Kidd metrics. CK metrics is one of the oldest and most reliable metrics among all metrics available to software industry to evaluate OO design. This paper presents an evaluation of CK metrics to propose an improved CK design metrics values to reduce the defects during software design phase in software. This paper will also describe that whether a significant effect of any CK design metrics exists on total number of defects per module or not. This is achieved by conducting survey in two software development companies. |
2004.02002 | Tiancheng Zhao | Tianchang Zhao and Kyusong Lee | Talk to Papers: Bringing Neural Question Answering to Academic Search | demo paper accepted at ACL 2020 | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce Talk to Papers, which exploits the recent open-domain question
answering (QA) techniques to improve the current experience of academic search.
It's designed to enable researchers to use natural language queries to find
precise answers and extract insights from a massive amount of academic papers.
We present a large improvement over classic search engine baseline on several
standard QA datasets and provide the community a collaborative data collection
tool to curate the first natural language processing research QA dataset via a
community effort.
| [
{
"created": "Sat, 4 Apr 2020 19:19:55 GMT",
"version": "v1"
},
{
"created": "Mon, 13 Apr 2020 14:38:11 GMT",
"version": "v2"
},
{
"created": "Thu, 21 May 2020 20:26:28 GMT",
"version": "v3"
}
] | 2020-05-25 | [
[
"Zhao",
"Tianchang",
""
],
[
"Lee",
"Kyusong",
""
]
] | We introduce Talk to Papers, which exploits the recent open-domain question answering (QA) techniques to improve the current experience of academic search. It's designed to enable researchers to use natural language queries to find precise answers and extract insights from a massive amount of academic papers. We present a large improvement over classic search engine baseline on several standard QA datasets and provide the community a collaborative data collection tool to curate the first natural language processing research QA dataset via a community effort. |
2312.03322 | Zhimiao Yu | Zhimiao Yu, Tiancheng Lin, Yi Xu | Background Clustering Pre-training for Few-shot Segmentation | 6 pages, 2 figures, ICIP 2023 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent few-shot segmentation (FSS) methods introduce an extra pre-training
stage before meta-training to obtain a stronger backbone, which has become a
standard step in few-shot learning. Despite the effectiveness, current
pre-training scheme suffers from the merged background problem: only base
classes are labelled as foregrounds, making it hard to distinguish between
novel classes and actual background. In this paper, we propose a new
pre-training scheme for FSS via decoupling the novel classes from background,
called Background Clustering Pre-Training (BCPT). Specifically, we adopt online
clustering to the pixel embeddings of merged background to explore the
underlying semantic structures, bridging the gap between pre-training and
adaptation to novel classes. Given the clustering results, we further propose
the background mining loss and leverage base classes to guide the clustering
process, improving the quality and stability of clustering results. Experiments
on PASCAL-5i and COCO-20i show that BCPT yields advanced performance. Code will
be available.
| [
{
"created": "Wed, 6 Dec 2023 07:16:32 GMT",
"version": "v1"
}
] | 2023-12-07 | [
[
"Yu",
"Zhimiao",
""
],
[
"Lin",
"Tiancheng",
""
],
[
"Xu",
"Yi",
""
]
] | Recent few-shot segmentation (FSS) methods introduce an extra pre-training stage before meta-training to obtain a stronger backbone, which has become a standard step in few-shot learning. Despite the effectiveness, current pre-training scheme suffers from the merged background problem: only base classes are labelled as foregrounds, making it hard to distinguish between novel classes and actual background. In this paper, we propose a new pre-training scheme for FSS via decoupling the novel classes from background, called Background Clustering Pre-Training (BCPT). Specifically, we adopt online clustering to the pixel embeddings of merged background to explore the underlying semantic structures, bridging the gap between pre-training and adaptation to novel classes. Given the clustering results, we further propose the background mining loss and leverage base classes to guide the clustering process, improving the quality and stability of clustering results. Experiments on PASCAL-5i and COCO-20i show that BCPT yields advanced performance. Code will be available. |
1609.03355 | Zhou Zhou | Zhou Zhou, Jun Fang, Linxiao Yang, Hongbin Li, Zhi Chen and Rick S.
Blum | Low-Rank Tensor Decomposition-Aided Channel Estimation for Millimeter
Wave MIMO-OFDM Systems | arXiv admin note: text overlap with arXiv:1602.07955 | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We consider the problem of downlink channel estimation for millimeter wave
(mmWave) MIMO-OFDM systems, where both the base station (BS) and the mobile
station (MS) employ large antenna arrays for directional precoding/beamforming.
Hybrid analog and digital beamforming structures are employed in order to offer
a compromise between hardware complexity and system performance. Different from
most existing studies that are concerned with narrowband channels, we consider
estimation of wideband mmWave channels with frequency selectivity, which is
more appropriate for mmWave MIMO-OFDM systems. By exploiting the sparse
scattering nature of mmWave channels, we propose a CANDECOMP/PARAFAC (CP)
decomposition-based method for channel parameter estimation (including angles
of arrival/departure, time delays, and fading coefficients). In our proposed
method, the received signal at the BS is expressed as a third-order tensor. We
show that the tensor has the form of a low-rank CP decomposition, and the
channel parameters can be estimated from the associated factor matrices. Our
analysis reveals that the uniqueness of the CP decomposition can be guaranteed
even when the size of the tensor is small. Hence the proposed method has the
potential to achieve substantial training overhead reduction. We also develop
Cramer-Rao bound (CRB) results for channel parameters, and compare our proposed
method with a compressed sensing-based method. Simulation results show that the
proposed method attains mean square errors that are very close to their
associated CRBs, and presents a clear advantage over the compressed
sensing-based method in terms of both estimation accuracy and computational
complexity.
| [
{
"created": "Mon, 12 Sep 2016 11:52:48 GMT",
"version": "v1"
},
{
"created": "Tue, 1 Nov 2016 08:45:05 GMT",
"version": "v2"
}
] | 2016-11-02 | [
[
"Zhou",
"Zhou",
""
],
[
"Fang",
"Jun",
""
],
[
"Yang",
"Linxiao",
""
],
[
"Li",
"Hongbin",
""
],
[
"Chen",
"Zhi",
""
],
[
"Blum",
"Rick S.",
""
]
] | We consider the problem of downlink channel estimation for millimeter wave (mmWave) MIMO-OFDM systems, where both the base station (BS) and the mobile station (MS) employ large antenna arrays for directional precoding/beamforming. Hybrid analog and digital beamforming structures are employed in order to offer a compromise between hardware complexity and system performance. Different from most existing studies that are concerned with narrowband channels, we consider estimation of wideband mmWave channels with frequency selectivity, which is more appropriate for mmWave MIMO-OFDM systems. By exploiting the sparse scattering nature of mmWave channels, we propose a CANDECOMP/PARAFAC (CP) decomposition-based method for channel parameter estimation (including angles of arrival/departure, time delays, and fading coefficients). In our proposed method, the received signal at the BS is expressed as a third-order tensor. We show that the tensor has the form of a low-rank CP decomposition, and the channel parameters can be estimated from the associated factor matrices. Our analysis reveals that the uniqueness of the CP decomposition can be guaranteed even when the size of the tensor is small. Hence the proposed method has the potential to achieve substantial training overhead reduction. We also develop Cramer-Rao bound (CRB) results for channel parameters, and compare our proposed method with a compressed sensing-based method. Simulation results show that the proposed method attains mean square errors that are very close to their associated CRBs, and presents a clear advantage over the compressed sensing-based method in terms of both estimation accuracy and computational complexity. |
2304.07825 | Samuel Epstein | Samuel Epstein | Regression and Algorithmic Information Theory | null | null | null | null | cs.CC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we prove a theorem about regression, in that the shortest
description of a function consistent with a finite sample of data is less than
the combined conditional Kolmogorov complexities over the data in the sample.
| [
{
"created": "Sun, 16 Apr 2023 16:30:38 GMT",
"version": "v1"
}
] | 2023-04-18 | [
[
"Epstein",
"Samuel",
""
]
] | In this paper we prove a theorem about regression, in that the shortest description of a function consistent with a finite sample of data is less than the combined conditional Kolmogorov complexities over the data in the sample. |
2209.01641 | Souvik Datta | Souvik Datta, Mangolik Kundu, Ratnadeep Das Choudhury, Sriramalakshmi
P, Sreedevi VT | IoT Book Bot | 2022 IEEE India Council International Subsections Conference
(INDISCON) | null | 10.1109/INDISCON54605.2022.9862937 | null | cs.HC cs.RO | http://creativecommons.org/licenses/by/4.0/ | In order to ease the process of library management many technologies have
been adopted but most of them focus on inventory management. There has hardly
been any progress of automation in the field of issuing and returning books to
the library on time. In colleges and schools, hostellers often forget to timely
return the issued books back to the library. To solve the above issue and to
ensure timely submission of the issued books, this work develops a Book-Bot
which solves these complexities. The bot can commute from point A to point B,
scan and verify QR Codes and Barcodes. The bot will have a certain payload
capacity for carrying books. The QR code and Barcode scanning will be enabled
by a Pi Camera, OpenCV and Raspberry Pi, thus making the exchange of books safe
and secure. The odometry maneuvers of the bot will be controlled manually via a
Blynk App. This paper focuses on how human intervention can be reduced and
automates the issue part of library management system with the help of a bot.
| [
{
"created": "Sun, 4 Sep 2022 15:30:17 GMT",
"version": "v1"
}
] | 2022-09-07 | [
[
"Datta",
"Souvik",
""
],
[
"Kundu",
"Mangolik",
""
],
[
"Choudhury",
"Ratnadeep Das",
""
],
[
"P",
"Sriramalakshmi",
""
],
[
"VT",
"Sreedevi",
""
]
] | In order to ease the process of library management many technologies have been adopted but most of them focus on inventory management. There has hardly been any progress of automation in the field of issuing and returning books to the library on time. In colleges and schools, hostellers often forget to timely return the issued books back to the library. To solve the above issue and to ensure timely submission of the issued books, this work develops a Book-Bot which solves these complexities. The bot can commute from point A to point B, scan and verify QR Codes and Barcodes. The bot will have a certain payload capacity for carrying books. The QR code and Barcode scanning will be enabled by a Pi Camera, OpenCV and Raspberry Pi, thus making the exchange of books safe and secure. The odometry maneuvers of the bot will be controlled manually via a Blynk App. This paper focuses on how human intervention can be reduced and automates the issue part of library management system with the help of a bot. |
2407.19199 | Ryosuke Motegi | Ryosuke Motegi and Yoichi Seki | A simulation study of cluster search algorithms in data set generated by
Gaussian mixture models | null | null | null | null | cs.LG stat.ML | http://creativecommons.org/licenses/by/4.0/ | Determining the number of clusters is a fundamental issue in data clustering.
Several algorithms have been proposed, including centroid-based algorithms
using the Euclidean distance and model-based algorithms using a mixture of
probability distributions. Among these, greedy algorithms for searching the
number of clusters by repeatedly splitting or merging clusters have advantages
in terms of computation time for problems with large sample sizes. However,
studies comparing these methods in systematic evaluation experiments still need
to be included. This study examines centroid- and model-based cluster search
algorithms in various cases that Gaussian mixture models (GMMs) can generate.
The cases are generated by combining five factors: dimensionality, sample size,
the number of clusters, cluster overlap, and covariance type. The results show
that some cluster-splitting criteria based on Euclidean distance make
unreasonable decisions when clusters overlap. The results also show that
model-based algorithms are insensitive to covariance type and cluster overlap
compared to the centroid-based method if the sample size is sufficient. Our
cluster search implementation codes are available at
https://github.com/lipryou/searchClustK
| [
{
"created": "Sat, 27 Jul 2024 07:47:25 GMT",
"version": "v1"
}
] | 2024-07-30 | [
[
"Motegi",
"Ryosuke",
""
],
[
"Seki",
"Yoichi",
""
]
] | Determining the number of clusters is a fundamental issue in data clustering. Several algorithms have been proposed, including centroid-based algorithms using the Euclidean distance and model-based algorithms using a mixture of probability distributions. Among these, greedy algorithms for searching the number of clusters by repeatedly splitting or merging clusters have advantages in terms of computation time for problems with large sample sizes. However, studies comparing these methods in systematic evaluation experiments still need to be included. This study examines centroid- and model-based cluster search algorithms in various cases that Gaussian mixture models (GMMs) can generate. The cases are generated by combining five factors: dimensionality, sample size, the number of clusters, cluster overlap, and covariance type. The results show that some cluster-splitting criteria based on Euclidean distance make unreasonable decisions when clusters overlap. The results also show that model-based algorithms are insensitive to covariance type and cluster overlap compared to the centroid-based method if the sample size is sufficient. Our cluster search implementation codes are available at https://github.com/lipryou/searchClustK |
2302.12987 | Yi Gao | Yi Gao, Miao Xu, Min-Ling Zhang | Complementary to Multiple Labels: A Correlation-Aware Correction
Approach | null | null | 10.1109/TPAMI.2024.3416384 | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | \textit{Complementary label learning} (CLL) requires annotators to give
\emph{irrelevant} labels instead of relevant labels for instances. Currently,
CLL has shown its promising performance on multi-class data by estimating a
transition matrix. However, current multi-class CLL techniques cannot work well
on multi-labeled data since they assume each instance is associated with one
label while each multi-labeled instance is relevant to multiple labels. Here,
we show theoretically how the estimated transition matrix in multi-class CLL
could be distorted in multi-labeled cases as they ignore co-existing relevant
labels. Moreover, theoretical findings reveal that calculating a transition
matrix from label correlations in \textit{multi-labeled CLL} (ML-CLL) needs
multi-labeled data, while this is unavailable for ML-CLL. To solve this issue,
we propose a two-step method to estimate the transition matrix from candidate
labels. Specifically, we first estimate an initial transition matrix by
decomposing the multi-label problem into a series of binary classification
problems, then the initial transition matrix is corrected by label correlations
to enforce the addition of relationships among labels. We further show that the
proposal is classifier-consistent, and additionally introduce an MSE-based
regularizer to alleviate the tendency of BCE loss overfitting to noises.
Experimental results have demonstrated the effectiveness of the proposed
method.
| [
{
"created": "Sat, 25 Feb 2023 04:48:48 GMT",
"version": "v1"
}
] | 2024-06-25 | [
[
"Gao",
"Yi",
""
],
[
"Xu",
"Miao",
""
],
[
"Zhang",
"Min-Ling",
""
]
] | \textit{Complementary label learning} (CLL) requires annotators to give \emph{irrelevant} labels instead of relevant labels for instances. Currently, CLL has shown its promising performance on multi-class data by estimating a transition matrix. However, current multi-class CLL techniques cannot work well on multi-labeled data since they assume each instance is associated with one label while each multi-labeled instance is relevant to multiple labels. Here, we show theoretically how the estimated transition matrix in multi-class CLL could be distorted in multi-labeled cases as they ignore co-existing relevant labels. Moreover, theoretical findings reveal that calculating a transition matrix from label correlations in \textit{multi-labeled CLL} (ML-CLL) needs multi-labeled data, while this is unavailable for ML-CLL. To solve this issue, we propose a two-step method to estimate the transition matrix from candidate labels. Specifically, we first estimate an initial transition matrix by decomposing the multi-label problem into a series of binary classification problems, then the initial transition matrix is corrected by label correlations to enforce the addition of relationships among labels. We further show that the proposal is classifier-consistent, and additionally introduce an MSE-based regularizer to alleviate the tendency of BCE loss overfitting to noises. Experimental results have demonstrated the effectiveness of the proposed method. |
2306.03894 | Todd Schmid | Todd Schmid and Victoria Noquez and Lawrence S. Moss | Fractals from Regular Behaviours | Expanded and edited into a journal version. Submitted to the CALCO
2023 special issue of LMCS. (31 pages, 5 figures.) | null | null | null | cs.LO cs.FL | http://creativecommons.org/licenses/by/4.0/ | We are interested in connections between the theory of fractal sets obtained
as attractors of iterated function systems and process calculi. To this end, we
reinterpret Milner's expressions for processes as contraction operators on a
complete metric space. When the space is, for example, the plane, the
denotations of fixed point terms correspond to familiar fractal sets. We give a
sound and complete axiomatization of fractal equivalence, the congruence on
terms consisting of pairs that construct identical self-similar sets in all
interpretations. We further make connections to labelled Markov chains and to
invariant measures. In all of this work, we use important results from process
calculi. For example, we use Rabinovich's completeness theorem for trace
equivalence in our own completeness theorem. In addition to our results, we
also raise several questions related to both fractals and process calculi.
| [
{
"created": "Tue, 6 Jun 2023 17:55:12 GMT",
"version": "v1"
},
{
"created": "Thu, 1 Feb 2024 16:18:26 GMT",
"version": "v2"
}
] | 2024-02-02 | [
[
"Schmid",
"Todd",
""
],
[
"Noquez",
"Victoria",
""
],
[
"Moss",
"Lawrence S.",
""
]
] | We are interested in connections between the theory of fractal sets obtained as attractors of iterated function systems and process calculi. To this end, we reinterpret Milner's expressions for processes as contraction operators on a complete metric space. When the space is, for example, the plane, the denotations of fixed point terms correspond to familiar fractal sets. We give a sound and complete axiomatization of fractal equivalence, the congruence on terms consisting of pairs that construct identical self-similar sets in all interpretations. We further make connections to labelled Markov chains and to invariant measures. In all of this work, we use important results from process calculi. For example, we use Rabinovich's completeness theorem for trace equivalence in our own completeness theorem. In addition to our results, we also raise several questions related to both fractals and process calculi. |
2208.04347 | Jason Phang | Jason Phang, Yao Zhao, Peter J. Liu | Investigating Efficiently Extending Transformers for Long Input
Summarization | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | While large pretrained Transformer models have proven highly capable at
tackling natural language tasks, handling long sequence inputs continues to be
a significant challenge. One such task is long input summarization, where
inputs are longer than the maximum input context of most pretrained models.
Through an extensive set of experiments, we investigate what model
architectural changes and pretraining paradigms can most efficiently adapt a
pretrained Transformer for long input summarization. We find that a staggered,
block-local Transformer with global encoder tokens strikes a good balance of
performance and efficiency, and that an additional pretraining phase on long
sequences meaningfully improves downstream summarization performance. Based on
our findings, we introduce PEGASUS-X, an extension of the PEGASUS model with
additional long input pretraining to handle inputs of up to 16K tokens.
PEGASUS-X achieves strong performance on long input summarization tasks
comparable with much larger models while adding few additional parameters and
not requiring model parallelism to train.
| [
{
"created": "Mon, 8 Aug 2022 18:10:58 GMT",
"version": "v1"
}
] | 2022-08-10 | [
[
"Phang",
"Jason",
""
],
[
"Zhao",
"Yao",
""
],
[
"Liu",
"Peter J.",
""
]
] | While large pretrained Transformer models have proven highly capable at tackling natural language tasks, handling long sequence inputs continues to be a significant challenge. One such task is long input summarization, where inputs are longer than the maximum input context of most pretrained models. Through an extensive set of experiments, we investigate what model architectural changes and pretraining paradigms can most efficiently adapt a pretrained Transformer for long input summarization. We find that a staggered, block-local Transformer with global encoder tokens strikes a good balance of performance and efficiency, and that an additional pretraining phase on long sequences meaningfully improves downstream summarization performance. Based on our findings, we introduce PEGASUS-X, an extension of the PEGASUS model with additional long input pretraining to handle inputs of up to 16K tokens. PEGASUS-X achieves strong performance on long input summarization tasks comparable with much larger models while adding few additional parameters and not requiring model parallelism to train. |
1404.5144 | Sacha Gomez | M. Konomi and G. M. Sacha | Influence of the learning method in the performance of feedforward
neural networks when the activity of neurons is modified | null | null | null | null | cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A method that allows us to give a different treatment to any neuron inside
feedforward neural networks is presented. The algorithm has been implemented
with two very different learning methods: a standard Back-propagation (BP)
procedure and an evolutionary algorithm. First, we have demonstrated that the
EA training method converges faster and gives more accurate results than BP.
Then we have made a full analysis of the effects of turning off different
combinations of neurons after the training phase. We demonstrate that EA is
much more robust than BP for all the cases under study. Even in the case when
two hidden neurons are lost, EA training is still able to give good average
results. This difference implies that we must be very careful when pruning or
redundancy effects are being studied since the network performance when losing
neurons strongly depends on the training method. Moreover, the influence of the
individual inputs will also depend on the training algorithm. Since EA keeps a
good classification performance when units are lost, this method could be a
good way to simulate biological learning systems since they must be robust
against deficient neuron performance. Although biological systems are much more
complex than the simulations shown in this article, we propose that a smart
training strategy such as the one shown here could be considered as a first
protection against the losing of a certain number of neurons.
| [
{
"created": "Mon, 21 Apr 2014 09:00:19 GMT",
"version": "v1"
}
] | 2014-04-22 | [
[
"Konomi",
"M.",
""
],
[
"Sacha",
"G. M.",
""
]
] | A method that allows us to give a different treatment to any neuron inside feedforward neural networks is presented. The algorithm has been implemented with two very different learning methods: a standard Back-propagation (BP) procedure and an evolutionary algorithm. First, we have demonstrated that the EA training method converges faster and gives more accurate results than BP. Then we have made a full analysis of the effects of turning off different combinations of neurons after the training phase. We demonstrate that EA is much more robust than BP for all the cases under study. Even in the case when two hidden neurons are lost, EA training is still able to give good average results. This difference implies that we must be very careful when pruning or redundancy effects are being studied since the network performance when losing neurons strongly depends on the training method. Moreover, the influence of the individual inputs will also depend on the training algorithm. Since EA keeps a good classification performance when units are lost, this method could be a good way to simulate biological learning systems since they must be robust against deficient neuron performance. Although biological systems are much more complex than the simulations shown in this article, we propose that a smart training strategy such as the one shown here could be considered as a first protection against the losing of a certain number of neurons. |
2407.00142 | Christopher Irwin | Christopher Irwin, Flavio Mignone, Stefania Montani, Luigi Portinale | Graph Neural Networks for Gut Microbiome Metaomic data: A preliminary
work | null | null | null | null | cs.LG cs.AI | http://creativecommons.org/licenses/by/4.0/ | The gut microbiome, crucial for human health, presents challenges in
analyzing its complex metaomic data due to high dimensionality and sparsity.
Traditional methods struggle to capture its intricate relationships. We
investigate graph neural networks (GNNs) for this task, aiming to derive
meaningful representations of individual gut microbiomes. Unlike methods
relying solely on taxa abundance, we directly leverage phylogenetic
relationships, in order to obtain a generalized encoder for taxa networks. The
representation learnt from the encoder are then used to train a model for
phenotype prediction such as Inflammatory Bowel Disease (IBD).
| [
{
"created": "Fri, 28 Jun 2024 15:53:36 GMT",
"version": "v1"
}
] | 2024-07-02 | [
[
"Irwin",
"Christopher",
""
],
[
"Mignone",
"Flavio",
""
],
[
"Montani",
"Stefania",
""
],
[
"Portinale",
"Luigi",
""
]
] | The gut microbiome, crucial for human health, presents challenges in analyzing its complex metaomic data due to high dimensionality and sparsity. Traditional methods struggle to capture its intricate relationships. We investigate graph neural networks (GNNs) for this task, aiming to derive meaningful representations of individual gut microbiomes. Unlike methods relying solely on taxa abundance, we directly leverage phylogenetic relationships, in order to obtain a generalized encoder for taxa networks. The representation learnt from the encoder are then used to train a model for phenotype prediction such as Inflammatory Bowel Disease (IBD). |
2402.11835 | Hugh Zhang | Luca D'Amico-Wong, Hugh Zhang, Marc Lanctot, David C. Parkes | Easy as ABCs: Unifying Boltzmann Q-Learning and Counterfactual Regret
Minimization | null | null | null | null | cs.LG cs.GT cs.MA | http://creativecommons.org/licenses/by/4.0/ | We propose ABCs (Adaptive Branching through Child stationarity), a
best-of-both-worlds algorithm combining Boltzmann Q-learning (BQL), a classic
reinforcement learning algorithm for single-agent domains, and counterfactual
regret minimization (CFR), a central algorithm for learning in multi-agent
domains. ABCs adaptively chooses what fraction of the environment to explore
each iteration by measuring the stationarity of the environment's reward and
transition dynamics. In Markov decision processes, ABCs converges to the
optimal policy with at most an O(A) factor slowdown compared to BQL, where A is
the number of actions in the environment. In two-player zero-sum games, ABCs is
guaranteed to converge to a Nash equilibrium (assuming access to a perfect
oracle for detecting stationarity), while BQL has no such guarantees.
Empirically, ABCs demonstrates strong performance when benchmarked across
environments drawn from the OpenSpiel game library and OpenAI Gym and exceeds
all prior methods in environments which are neither fully stationary nor fully
nonstationary.
| [
{
"created": "Mon, 19 Feb 2024 04:58:39 GMT",
"version": "v1"
}
] | 2024-02-20 | [
[
"D'Amico-Wong",
"Luca",
""
],
[
"Zhang",
"Hugh",
""
],
[
"Lanctot",
"Marc",
""
],
[
"Parkes",
"David C.",
""
]
] | We propose ABCs (Adaptive Branching through Child stationarity), a best-of-both-worlds algorithm combining Boltzmann Q-learning (BQL), a classic reinforcement learning algorithm for single-agent domains, and counterfactual regret minimization (CFR), a central algorithm for learning in multi-agent domains. ABCs adaptively chooses what fraction of the environment to explore each iteration by measuring the stationarity of the environment's reward and transition dynamics. In Markov decision processes, ABCs converges to the optimal policy with at most an O(A) factor slowdown compared to BQL, where A is the number of actions in the environment. In two-player zero-sum games, ABCs is guaranteed to converge to a Nash equilibrium (assuming access to a perfect oracle for detecting stationarity), while BQL has no such guarantees. Empirically, ABCs demonstrates strong performance when benchmarked across environments drawn from the OpenSpiel game library and OpenAI Gym and exceeds all prior methods in environments which are neither fully stationary nor fully nonstationary. |
2307.01646 | Qi Yan | Qi Yan, Zhengyang Liang, Yang Song, Renjie Liao, Lele Wang | SwinGNN: Rethinking Permutation Invariance in Diffusion Models for Graph
Generation | TMLR 2024 | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion models based on permutation-equivariant networks can learn
permutation-invariant distributions for graph data. However, in comparison to
their non-invariant counterparts, we have found that these invariant models
encounter greater learning challenges since 1) their effective target
distributions exhibit more modes; 2) their optimal one-step denoising scores
are the score functions of Gaussian mixtures with more components. Motivated by
this analysis, we propose a non-invariant diffusion model, called
$\textit{SwinGNN}$, which employs an efficient edge-to-edge 2-WL message
passing network and utilizes shifted window based self-attention inspired by
SwinTransformers. Further, through systematic ablations, we identify several
critical training and sampling techniques that significantly improve the sample
quality of graph generation. At last, we introduce a simple post-processing
trick, $\textit{i.e.}$, randomly permuting the generated graphs, which provably
converts any graph generative model to a permutation-invariant one. Extensive
experiments on synthetic and real-world protein and molecule datasets show that
our SwinGNN achieves state-of-the-art performances. Our code is released at
https://github.com/qiyan98/SwinGNN.
| [
{
"created": "Tue, 4 Jul 2023 10:58:42 GMT",
"version": "v1"
},
{
"created": "Wed, 19 Jul 2023 04:59:35 GMT",
"version": "v2"
},
{
"created": "Tue, 18 Jun 2024 05:55:32 GMT",
"version": "v3"
},
{
"created": "Wed, 19 Jun 2024 04:48:13 GMT",
"version": "v4"
}
] | 2024-06-21 | [
[
"Yan",
"Qi",
""
],
[
"Liang",
"Zhengyang",
""
],
[
"Song",
"Yang",
""
],
[
"Liao",
"Renjie",
""
],
[
"Wang",
"Lele",
""
]
] | Diffusion models based on permutation-equivariant networks can learn permutation-invariant distributions for graph data. However, in comparison to their non-invariant counterparts, we have found that these invariant models encounter greater learning challenges since 1) their effective target distributions exhibit more modes; 2) their optimal one-step denoising scores are the score functions of Gaussian mixtures with more components. Motivated by this analysis, we propose a non-invariant diffusion model, called $\textit{SwinGNN}$, which employs an efficient edge-to-edge 2-WL message passing network and utilizes shifted window based self-attention inspired by SwinTransformers. Further, through systematic ablations, we identify several critical training and sampling techniques that significantly improve the sample quality of graph generation. At last, we introduce a simple post-processing trick, $\textit{i.e.}$, randomly permuting the generated graphs, which provably converts any graph generative model to a permutation-invariant one. Extensive experiments on synthetic and real-world protein and molecule datasets show that our SwinGNN achieves state-of-the-art performances. Our code is released at https://github.com/qiyan98/SwinGNN. |
1706.07535 | Hemanth Venkateswara | Hemanth Venkateswara, Prasanth Lade, Binbin Lin, Jieping Ye,
Sethuraman Panchanathan | Efficient Approximate Solutions to Mutual Information Based Global
Feature Selection | ICDM 2015 Conference | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Mutual Information (MI) is often used for feature selection when developing
classifier models. Estimating the MI for a subset of features is often
intractable. We demonstrate, that under the assumptions of conditional
independence, MI between a subset of features can be expressed as the
Conditional Mutual Information (CMI) between pairs of features. But selecting
features with the highest CMI turns out to be a hard combinatorial problem. In
this work, we have applied two unique global methods, Truncated Power Method
(TPower) and Low Rank Bilinear Approximation (LowRank), to solve the feature
selection problem. These algorithms provide very good approximations to the
NP-hard CMI based feature selection problem. We experimentally demonstrate the
effectiveness of these procedures across multiple datasets and compare them
with existing MI based global and iterative feature selection procedures.
| [
{
"created": "Fri, 23 Jun 2017 01:08:59 GMT",
"version": "v1"
}
] | 2017-06-26 | [
[
"Venkateswara",
"Hemanth",
""
],
[
"Lade",
"Prasanth",
""
],
[
"Lin",
"Binbin",
""
],
[
"Ye",
"Jieping",
""
],
[
"Panchanathan",
"Sethuraman",
""
]
] | Mutual Information (MI) is often used for feature selection when developing classifier models. Estimating the MI for a subset of features is often intractable. We demonstrate, that under the assumptions of conditional independence, MI between a subset of features can be expressed as the Conditional Mutual Information (CMI) between pairs of features. But selecting features with the highest CMI turns out to be a hard combinatorial problem. In this work, we have applied two unique global methods, Truncated Power Method (TPower) and Low Rank Bilinear Approximation (LowRank), to solve the feature selection problem. These algorithms provide very good approximations to the NP-hard CMI based feature selection problem. We experimentally demonstrate the effectiveness of these procedures across multiple datasets and compare them with existing MI based global and iterative feature selection procedures. |
2109.05234 | Zezhong Wang Mr. | Zezhong Wang, Hongru Wang, Kwan Wai Chung, Jia Zhu, Gabriel Pui Cheong
Fung, Kam-Fai Wong | Prior Omission of Dissimilar Source Domain(s) for Cost-Effective
Few-Shot Learning | null | null | null | null | cs.CL cs.AI | http://creativecommons.org/licenses/by-nc-nd/4.0/ | Few-shot slot tagging is an emerging research topic in the field of Natural
Language Understanding (NLU). With sufficient annotated data from source
domains, the key challenge is how to train and adapt the model to another
target domain which only has few labels. Conventional few-shot approaches use
all the data from the source domains without considering inter-domain relations
and implicitly assume each sample in the domain contributes equally. However,
our experiments show that the data distribution bias among different domains
will significantly affect the adaption performance. Moreover, transferring
knowledge from dissimilar domains will even introduce some extra noises so that
affect the performance of models. To tackle this problem, we propose an
effective similarity-based method to select data from the source domains. In
addition, we propose a Shared-Private Network (SP-Net) for the few-shot slot
tagging task. The words from the same class would have some shared features. We
extract those shared features from the limited annotated data on the target
domain and merge them together as the label embedding to help us predict other
unlabelled data on the target domain. The experiment shows that our method
outperforms the state-of-the-art approaches with fewer source data. The result
also proves that some training data from dissimilar sources are redundant and
even negative for the adaption.
| [
{
"created": "Sat, 11 Sep 2021 09:30:59 GMT",
"version": "v1"
}
] | 2021-09-14 | [
[
"Wang",
"Zezhong",
""
],
[
"Wang",
"Hongru",
""
],
[
"Chung",
"Kwan Wai",
""
],
[
"Zhu",
"Jia",
""
],
[
"Fung",
"Gabriel Pui Cheong",
""
],
[
"Wong",
"Kam-Fai",
""
]
] | Few-shot slot tagging is an emerging research topic in the field of Natural Language Understanding (NLU). With sufficient annotated data from source domains, the key challenge is how to train and adapt the model to another target domain which only has few labels. Conventional few-shot approaches use all the data from the source domains without considering inter-domain relations and implicitly assume each sample in the domain contributes equally. However, our experiments show that the data distribution bias among different domains will significantly affect the adaption performance. Moreover, transferring knowledge from dissimilar domains will even introduce some extra noises so that affect the performance of models. To tackle this problem, we propose an effective similarity-based method to select data from the source domains. In addition, we propose a Shared-Private Network (SP-Net) for the few-shot slot tagging task. The words from the same class would have some shared features. We extract those shared features from the limited annotated data on the target domain and merge them together as the label embedding to help us predict other unlabelled data on the target domain. The experiment shows that our method outperforms the state-of-the-art approaches with fewer source data. The result also proves that some training data from dissimilar sources are redundant and even negative for the adaption. |
2104.01040 | Igor Halperin | Igor Halperin | Distributional Offline Continuous-Time Reinforcement Learning with
Neural Physics-Informed PDEs (SciPhy RL for DOCTR-L) | 24 pages, 5 figures | null | null | null | cs.LG cs.AI physics.comp-ph q-fin.CP | http://creativecommons.org/licenses/by-nc-sa/4.0/ | This paper addresses distributional offline continuous-time reinforcement
learning (DOCTR-L) with stochastic policies for high-dimensional optimal
control. A soft distributional version of the classical Hamilton-Jacobi-Bellman
(HJB) equation is given by a semilinear partial differential equation (PDE).
This `soft HJB equation' can be learned from offline data without assuming that
the latter correspond to a previous optimal or near-optimal policy. A
data-driven solution of the soft HJB equation uses methods of Neural PDEs and
Physics-Informed Neural Networks developed in the field of Scientific Machine
Learning (SciML). The suggested approach, dubbed `SciPhy RL', thus reduces
DOCTR-L to solving neural PDEs from data. Our algorithm called Deep DOCTR-L
converts offline high-dimensional data into an optimal policy in one step by
reducing it to supervised learning, instead of relying on value iteration or
policy iteration methods. The method enables a computable approach to the
quality control of obtained policies in terms of both their expected returns
and uncertainties about their values.
| [
{
"created": "Fri, 2 Apr 2021 13:22:14 GMT",
"version": "v1"
}
] | 2021-04-05 | [
[
"Halperin",
"Igor",
""
]
] | This paper addresses distributional offline continuous-time reinforcement learning (DOCTR-L) with stochastic policies for high-dimensional optimal control. A soft distributional version of the classical Hamilton-Jacobi-Bellman (HJB) equation is given by a semilinear partial differential equation (PDE). This `soft HJB equation' can be learned from offline data without assuming that the latter correspond to a previous optimal or near-optimal policy. A data-driven solution of the soft HJB equation uses methods of Neural PDEs and Physics-Informed Neural Networks developed in the field of Scientific Machine Learning (SciML). The suggested approach, dubbed `SciPhy RL', thus reduces DOCTR-L to solving neural PDEs from data. Our algorithm called Deep DOCTR-L converts offline high-dimensional data into an optimal policy in one step by reducing it to supervised learning, instead of relying on value iteration or policy iteration methods. The method enables a computable approach to the quality control of obtained policies in terms of both their expected returns and uncertainties about their values. |
1006.2022 | Haim Permuter Henry | Haim Permuter, Shlomo (Shitz) Shamai, and Anelia Somekh-Baruch | Message and state cooperation in multiple access channels | null | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the capacity of a multiple access channel with cooperating
encoders where partial state information is known to each encoder and full
state information is known to the decoder. The cooperation between the encoders
has a two-fold purpose: to generate empirical state coordination between the
encoders, and to share information about the private messages that each encoder
has. For two-way cooperation, this two-fold purpose is achieved by
double-binning, where the first layer of binning is used to generate the state
coordination similarly to the two-way source coding, and the second layer of
binning is used to transmit information about the private messages. The
complete result provides the framework and perspective for addressing a complex
level of cooperation that mixes states and messages in an optimal way.
| [
{
"created": "Thu, 10 Jun 2010 13:15:44 GMT",
"version": "v1"
}
] | 2010-06-11 | [
[
"Permuter",
"Haim",
"",
"Shitz"
],
[
"Shlomo",
"",
"",
"Shitz"
],
[
"Shamai",
"",
""
],
[
"Somekh-Baruch",
"Anelia",
""
]
] | We investigate the capacity of a multiple access channel with cooperating encoders where partial state information is known to each encoder and full state information is known to the decoder. The cooperation between the encoders has a two-fold purpose: to generate empirical state coordination between the encoders, and to share information about the private messages that each encoder has. For two-way cooperation, this two-fold purpose is achieved by double-binning, where the first layer of binning is used to generate the state coordination similarly to the two-way source coding, and the second layer of binning is used to transmit information about the private messages. The complete result provides the framework and perspective for addressing a complex level of cooperation that mixes states and messages in an optimal way. |
2303.11899 | Hankang Gu | Hankang Gu, Shangbo Wang, Xiaoguang Ma, Dongyao Jia, Guoqiang Mao, Eng
Gee Lim, Cheuk Pong Ryan Wong | Large-Scale Traffic Signal Control Using Constrained Network Partition
and Adaptive Deep Reinforcement Learning | null | null | null | null | cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Multi-agent Deep Reinforcement Learning (MADRL) based traffic signal control
becomes a popular research topic in recent years. To alleviate the scalability
issue of completely centralized RL techniques and the non-stationarity issue of
completely decentralized RL techniques on large-scale traffic networks, some
literature utilizes a regional control approach where the whole network is
firstly partitioned into multiple disjoint regions, followed by applying the
centralized RL approach to each region. However, the existing partitioning
rules either have no constraints on the topology of regions or require the same
topology for all regions. Meanwhile, no existing regional control approach
explores the performance of optimal joint action in an exponentially growing
regional action space when intersections are controlled by 4-phase traffic
signals (EW, EWL, NS, NSL). In this paper, we propose a novel RL training
framework named RegionLight to tackle the above limitations. Specifically, the
topology of regions is firstly constrained to a star network which comprises
one center and an arbitrary number of leaves. Next, the network partitioning
problem is modeled as an optimization problem to minimize the number of
regions. Then, an Adaptive Branching Dueling Q-Network (ABDQ) model is proposed
to decompose the regional control task into several joint signal control
sub-tasks corresponding to particular intersections. Subsequently, these
sub-tasks maximize the regional benefits cooperatively. Finally, the global
control strategy for the whole network is obtained by concatenating the optimal
joint actions of all regions. Experimental results demonstrate the superiority
of our proposed framework over all baselines under both real and synthetic
datasets in all evaluation metrics.
| [
{
"created": "Tue, 21 Mar 2023 14:42:58 GMT",
"version": "v1"
},
{
"created": "Wed, 22 Mar 2023 07:34:22 GMT",
"version": "v2"
},
{
"created": "Fri, 7 Apr 2023 06:38:44 GMT",
"version": "v3"
},
{
"created": "Mon, 26 Jun 2023 04:08:48 GMT",
"version": "v4"
},
{
"created": "Thu, 7 Sep 2023 04:42:45 GMT",
"version": "v5"
}
] | 2023-09-08 | [
[
"Gu",
"Hankang",
""
],
[
"Wang",
"Shangbo",
""
],
[
"Ma",
"Xiaoguang",
""
],
[
"Jia",
"Dongyao",
""
],
[
"Mao",
"Guoqiang",
""
],
[
"Lim",
"Eng Gee",
""
],
[
"Wong",
"Cheuk Pong Ryan",
""
]
] | Multi-agent Deep Reinforcement Learning (MADRL) based traffic signal control becomes a popular research topic in recent years. To alleviate the scalability issue of completely centralized RL techniques and the non-stationarity issue of completely decentralized RL techniques on large-scale traffic networks, some literature utilizes a regional control approach where the whole network is firstly partitioned into multiple disjoint regions, followed by applying the centralized RL approach to each region. However, the existing partitioning rules either have no constraints on the topology of regions or require the same topology for all regions. Meanwhile, no existing regional control approach explores the performance of optimal joint action in an exponentially growing regional action space when intersections are controlled by 4-phase traffic signals (EW, EWL, NS, NSL). In this paper, we propose a novel RL training framework named RegionLight to tackle the above limitations. Specifically, the topology of regions is firstly constrained to a star network which comprises one center and an arbitrary number of leaves. Next, the network partitioning problem is modeled as an optimization problem to minimize the number of regions. Then, an Adaptive Branching Dueling Q-Network (ABDQ) model is proposed to decompose the regional control task into several joint signal control sub-tasks corresponding to particular intersections. Subsequently, these sub-tasks maximize the regional benefits cooperatively. Finally, the global control strategy for the whole network is obtained by concatenating the optimal joint actions of all regions. Experimental results demonstrate the superiority of our proposed framework over all baselines under both real and synthetic datasets in all evaluation metrics. |
2210.07806 | Luca Canalini | Luca Canalini, Jan Klein, Nuno Pedrosa de Barros, Diana Maria Sima,
Dorothea Miller, Horst Hahn | Comparison of different automatic solutions for resection cavity
segmentation in postoperative MRI volumes including longitudinal acquisitions | null | SPIE Proceedings Vol. 11598 - Medical Imaging 2021: Image-Guided
Procedures, Robotic Interventions, and Modeling | 10.1117/12.2580889 | null | cs.CV cs.AI | http://creativecommons.org/licenses/by/4.0/ | In this work, we compare five deep learning solutions to automatically
segment the resection cavity in postoperative MRI. The proposed methods are
based on the same 3D U-Net architecture. We use a dataset of postoperative MRI
volumes, each including four MRI sequences and the ground truth of the
corresponding resection cavity. Four solutions are trained with a different MRI
sequence. Besides, a method designed with all the available sequences is also
presented. Our experiments show that the method trained only with the T1
weighted contrast-enhanced MRI sequence achieves the best results, with a
median DICE index of 0.81.
| [
{
"created": "Fri, 14 Oct 2022 13:37:35 GMT",
"version": "v1"
}
] | 2022-10-17 | [
[
"Canalini",
"Luca",
""
],
[
"Klein",
"Jan",
""
],
[
"de Barros",
"Nuno Pedrosa",
""
],
[
"Sima",
"Diana Maria",
""
],
[
"Miller",
"Dorothea",
""
],
[
"Hahn",
"Horst",
""
]
] | In this work, we compare five deep learning solutions to automatically segment the resection cavity in postoperative MRI. The proposed methods are based on the same 3D U-Net architecture. We use a dataset of postoperative MRI volumes, each including four MRI sequences and the ground truth of the corresponding resection cavity. Four solutions are trained with a different MRI sequence. Besides, a method designed with all the available sequences is also presented. Our experiments show that the method trained only with the T1 weighted contrast-enhanced MRI sequence achieves the best results, with a median DICE index of 0.81. |
2405.03287 | Mehedi Hasan Raju | Mehedi Hasan Raju, Dillon J Lohr, Oleg V Komogortsev | Evaluating Eye Movement Biometrics in Virtual Reality: A Comparative
Analysis of VR Headset and High-End Eye-Tracker Collected Dataset | 9 pages, 6 figures | null | null | null | cs.HC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | Previous studies have shown that eye movement data recorded at 1000 Hz can be
used to authenticate individuals. This study explores the effectiveness of eye
movement-based biometrics (EMB) by utilizing data from an eye-tracking
(ET)-enabled virtual reality (VR) headset (GazeBaseVR) and compares it to the
performance using data from a high-end eye tracker (GazeBase) that has been
downsampled to 250 Hz. The research also aims to assess the biometric potential
of both binocular and monocular eye movement data. GazeBaseVR dataset achieves
an equal error rate (EER) of 1.67% and a false rejection rate (FRR) at 10^-4
false acceptance rate (FAR) of 22.73% in a binocular configuration. This study
underscores the biometric viability of data obtained from eye-tracking-enabled
VR headset.
| [
{
"created": "Mon, 6 May 2024 09:05:06 GMT",
"version": "v1"
}
] | 2024-05-07 | [
[
"Raju",
"Mehedi Hasan",
""
],
[
"Lohr",
"Dillon J",
""
],
[
"Komogortsev",
"Oleg V",
""
]
] | Previous studies have shown that eye movement data recorded at 1000 Hz can be used to authenticate individuals. This study explores the effectiveness of eye movement-based biometrics (EMB) by utilizing data from an eye-tracking (ET)-enabled virtual reality (VR) headset (GazeBaseVR) and compares it to the performance using data from a high-end eye tracker (GazeBase) that has been downsampled to 250 Hz. The research also aims to assess the biometric potential of both binocular and monocular eye movement data. GazeBaseVR dataset achieves an equal error rate (EER) of 1.67% and a false rejection rate (FRR) at 10^-4 false acceptance rate (FAR) of 22.73% in a binocular configuration. This study underscores the biometric viability of data obtained from eye-tracking-enabled VR headset. |
2211.14923 | Jiayu Song | Jiayu Song, Iman Munire Bilal, Adam Tsakalidis, Rob Procter, Maria
Liakata | Unsupervised Opinion Summarisation in the Wasserstein Space | null | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Opinion summarisation synthesises opinions expressed in a group of documents
discussing the same topic to produce a single summary. Recent work has looked
at opinion summarisation of clusters of social media posts. Such posts are
noisy and have unpredictable structure, posing additional challenges for the
construction of the summary distribution and the preservation of meaning
compared to online reviews, which has been so far the focus of opinion
summarisation. To address these challenges we present \textit{WassOS}, an
unsupervised abstractive summarization model which makes use of the Wasserstein
distance. A Variational Autoencoder is used to get the distribution of
documents/posts, and the distributions are disentangled into separate semantic
and syntactic spaces. The summary distribution is obtained using the
Wasserstein barycenter of the semantic and syntactic distributions. A latent
variable sampled from the summary distribution is fed into a GRU decoder with a
transformer layer to produce the final summary. Our experiments on multiple
datasets including Twitter clusters, Reddit threads, and reviews show that
WassOS almost always outperforms the state-of-the-art on ROUGE metrics and
consistently produces the best summaries with respect to meaning preservation
according to human evaluations.
| [
{
"created": "Sun, 27 Nov 2022 19:45:38 GMT",
"version": "v1"
}
] | 2022-11-29 | [
[
"Song",
"Jiayu",
""
],
[
"Bilal",
"Iman Munire",
""
],
[
"Tsakalidis",
"Adam",
""
],
[
"Procter",
"Rob",
""
],
[
"Liakata",
"Maria",
""
]
] | Opinion summarisation synthesises opinions expressed in a group of documents discussing the same topic to produce a single summary. Recent work has looked at opinion summarisation of clusters of social media posts. Such posts are noisy and have unpredictable structure, posing additional challenges for the construction of the summary distribution and the preservation of meaning compared to online reviews, which has been so far the focus of opinion summarisation. To address these challenges we present \textit{WassOS}, an unsupervised abstractive summarization model which makes use of the Wasserstein distance. A Variational Autoencoder is used to get the distribution of documents/posts, and the distributions are disentangled into separate semantic and syntactic spaces. The summary distribution is obtained using the Wasserstein barycenter of the semantic and syntactic distributions. A latent variable sampled from the summary distribution is fed into a GRU decoder with a transformer layer to produce the final summary. Our experiments on multiple datasets including Twitter clusters, Reddit threads, and reviews show that WassOS almost always outperforms the state-of-the-art on ROUGE metrics and consistently produces the best summaries with respect to meaning preservation according to human evaluations. |
2012.08489 | Valerio Perrone | Valerio Perrone, Huibin Shen, Aida Zolic, Iaroslav Shcherbatyi, Amr
Ahmed, Tanya Bansal, Michele Donini, Fela Winkelmolen, Rodolphe Jenatton,
Jean Baptiste Faddoul, Barbara Pogorzelska, Miroslav Miladinovic, Krishnaram
Kenthapadi, Matthias Seeger, C\'edric Archambeau | Amazon SageMaker Automatic Model Tuning: Scalable Gradient-Free
Optimization | null | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Tuning complex machine learning systems is challenging. Machine learning
typically requires to set hyperparameters, be it regularization, architecture,
or optimization parameters, whose tuning is critical to achieve good predictive
performance. To democratize access to machine learning systems, it is essential
to automate the tuning. This paper presents Amazon SageMaker Automatic Model
Tuning (AMT), a fully managed system for gradient-free optimization at scale.
AMT finds the best version of a trained machine learning model by repeatedly
evaluating it with different hyperparameter configurations. It leverages either
random search or Bayesian optimization to choose the hyperparameter values
resulting in the best model, as measured by the metric chosen by the user. AMT
can be used with built-in algorithms, custom algorithms, and Amazon SageMaker
pre-built containers for machine learning frameworks. We discuss the core
functionality, system architecture, our design principles, and lessons learned.
We also describe more advanced features of AMT, such as automated early
stopping and warm-starting, showing in experiments their benefits to users.
| [
{
"created": "Tue, 15 Dec 2020 18:34:34 GMT",
"version": "v1"
},
{
"created": "Fri, 18 Jun 2021 19:41:09 GMT",
"version": "v2"
}
] | 2021-06-22 | [
[
"Perrone",
"Valerio",
""
],
[
"Shen",
"Huibin",
""
],
[
"Zolic",
"Aida",
""
],
[
"Shcherbatyi",
"Iaroslav",
""
],
[
"Ahmed",
"Amr",
""
],
[
"Bansal",
"Tanya",
""
],
[
"Donini",
"Michele",
""
],
[
"Winkelmolen",
"Fela",
""
],
[
"Jenatton",
"Rodolphe",
""
],
[
"Faddoul",
"Jean Baptiste",
""
],
[
"Pogorzelska",
"Barbara",
""
],
[
"Miladinovic",
"Miroslav",
""
],
[
"Kenthapadi",
"Krishnaram",
""
],
[
"Seeger",
"Matthias",
""
],
[
"Archambeau",
"Cédric",
""
]
] | Tuning complex machine learning systems is challenging. Machine learning typically requires to set hyperparameters, be it regularization, architecture, or optimization parameters, whose tuning is critical to achieve good predictive performance. To democratize access to machine learning systems, it is essential to automate the tuning. This paper presents Amazon SageMaker Automatic Model Tuning (AMT), a fully managed system for gradient-free optimization at scale. AMT finds the best version of a trained machine learning model by repeatedly evaluating it with different hyperparameter configurations. It leverages either random search or Bayesian optimization to choose the hyperparameter values resulting in the best model, as measured by the metric chosen by the user. AMT can be used with built-in algorithms, custom algorithms, and Amazon SageMaker pre-built containers for machine learning frameworks. We discuss the core functionality, system architecture, our design principles, and lessons learned. We also describe more advanced features of AMT, such as automated early stopping and warm-starting, showing in experiments their benefits to users. |
1906.10724 | Rahul Aralikatte | Rahul Aralikatte and Anders S{\o}gaard | Model-based annotation of coreference | To appear in LREC 2020 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Humans do not make inferences over texts, but over models of what texts are
about. When annotators are asked to annotate coreferent spans of text, it is
therefore a somewhat unnatural task. This paper presents an alternative in
which we preprocess documents, linking entities to a knowledge base, and turn
the coreference annotation task -- in our case limited to pronouns -- into an
annotation task where annotators are asked to assign pronouns to entities.
Model-based annotation is shown to lead to faster annotation and higher
inter-annotator agreement, and we argue that it also opens up for an
alternative approach to coreference resolution. We present two new coreference
benchmark datasets, for English Wikipedia and English teacher-student
dialogues, and evaluate state-of-the-art coreference resolvers on them.
| [
{
"created": "Tue, 25 Jun 2019 18:56:36 GMT",
"version": "v1"
},
{
"created": "Fri, 30 Aug 2019 08:25:13 GMT",
"version": "v2"
},
{
"created": "Sun, 1 Mar 2020 23:17:56 GMT",
"version": "v3"
}
] | 2020-03-03 | [
[
"Aralikatte",
"Rahul",
""
],
[
"Søgaard",
"Anders",
""
]
] | Humans do not make inferences over texts, but over models of what texts are about. When annotators are asked to annotate coreferent spans of text, it is therefore a somewhat unnatural task. This paper presents an alternative in which we preprocess documents, linking entities to a knowledge base, and turn the coreference annotation task -- in our case limited to pronouns -- into an annotation task where annotators are asked to assign pronouns to entities. Model-based annotation is shown to lead to faster annotation and higher inter-annotator agreement, and we argue that it also opens up for an alternative approach to coreference resolution. We present two new coreference benchmark datasets, for English Wikipedia and English teacher-student dialogues, and evaluate state-of-the-art coreference resolvers on them. |
1609.07190 | Rishab Nithyanand | Narseo Vallina-Rodriguez, Srikanth Sundaresan, Abbas Razaghpanah,
Rishab Nithyanand, Mark Allman, Christian Kreibich, Phillipa Gill | Tracking the Trackers: Towards Understanding the Mobile Advertising and
Tracking Ecosystem | null | null | null | null | cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Third-party services form an integral part of the mobile ecosystem: they
allow app developers to add features such as performance analytics and social
network integration, and to monetize their apps by enabling user tracking and
targeted ad delivery. At present users, researchers, and regulators all have at
best limited understanding of this third-party ecosystem. In this paper we seek
to shrink this gap. Using data from users of our ICSI Haystack app we gain a
rich view of the mobile ecosystem: we identify and characterize domains
associated with mobile advertising and user tracking, thereby taking an
important step towards greater transparency. We furthermore outline our steps
towards a public catalog and census of analytics services, their behavior,
their personal data collection processes, and their use across mobile apps.
| [
{
"created": "Thu, 22 Sep 2016 23:45:20 GMT",
"version": "v1"
},
{
"created": "Wed, 26 Oct 2016 15:50:14 GMT",
"version": "v2"
}
] | 2016-10-27 | [
[
"Vallina-Rodriguez",
"Narseo",
""
],
[
"Sundaresan",
"Srikanth",
""
],
[
"Razaghpanah",
"Abbas",
""
],
[
"Nithyanand",
"Rishab",
""
],
[
"Allman",
"Mark",
""
],
[
"Kreibich",
"Christian",
""
],
[
"Gill",
"Phillipa",
""
]
] | Third-party services form an integral part of the mobile ecosystem: they allow app developers to add features such as performance analytics and social network integration, and to monetize their apps by enabling user tracking and targeted ad delivery. At present users, researchers, and regulators all have at best limited understanding of this third-party ecosystem. In this paper we seek to shrink this gap. Using data from users of our ICSI Haystack app we gain a rich view of the mobile ecosystem: we identify and characterize domains associated with mobile advertising and user tracking, thereby taking an important step towards greater transparency. We furthermore outline our steps towards a public catalog and census of analytics services, their behavior, their personal data collection processes, and their use across mobile apps. |
2201.03472 | Peter Nightingale | Peter Nightingale | Savile Row Manual | arXiv admin note: substantial text overlap with arXiv:1601.02865 | null | null | null | cs.AI | http://creativecommons.org/licenses/by-sa/4.0/ | We describe the constraint modelling tool Savile Row, its input language and
its main features. Savile Row translates a solver-independent constraint
modelling language to the input languages for various solvers including
constraint, SAT, and SMT solvers. After a brief introduction, the manual
describes the Essence Prime language, which is the input language of Savile
Row. Then we describe the functions of the tool, its main features and options
and how to install and use it.
| [
{
"created": "Fri, 12 Nov 2021 09:47:55 GMT",
"version": "v1"
},
{
"created": "Tue, 30 Jul 2024 13:31:56 GMT",
"version": "v2"
}
] | 2024-07-31 | [
[
"Nightingale",
"Peter",
""
]
] | We describe the constraint modelling tool Savile Row, its input language and its main features. Savile Row translates a solver-independent constraint modelling language to the input languages for various solvers including constraint, SAT, and SMT solvers. After a brief introduction, the manual describes the Essence Prime language, which is the input language of Savile Row. Then we describe the functions of the tool, its main features and options and how to install and use it. |
1802.00507 | Manuel Ortega-Rodr\'iguez | Diana Valverde-M\'endez, Manuel Ortega-Rodr\'iguez, Hugo
Sol\'is-S\'anchez, Ariadna Venegas-Li | The effects of anger on automated long-term-spectra based
speaker-identification | 11 pages | null | null | null | cs.HC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Forensic speaker identification has traditionally considered approaches based
on long term spectra analysis as especially robust, given that they work well
for short recordings, are not sensitive to changes in the intensity of the
sample, and continue to function in the presence of noise and limited passband.
We find, however, that anger induces a significant distortion of the acoustic
signal for long term spectra analysis purposes. Even moderate anger offsets
speaker identification results by 33% in the direction of a different speaker
altogether. Thus, caution should be exercised when applying this tool.
| [
{
"created": "Sat, 20 Jan 2018 03:51:30 GMT",
"version": "v1"
}
] | 2018-02-05 | [
[
"Valverde-Méndez",
"Diana",
""
],
[
"Ortega-Rodríguez",
"Manuel",
""
],
[
"Solís-Sánchez",
"Hugo",
""
],
[
"Venegas-Li",
"Ariadna",
""
]
] | Forensic speaker identification has traditionally considered approaches based on long term spectra analysis as especially robust, given that they work well for short recordings, are not sensitive to changes in the intensity of the sample, and continue to function in the presence of noise and limited passband. We find, however, that anger induces a significant distortion of the acoustic signal for long term spectra analysis purposes. Even moderate anger offsets speaker identification results by 33% in the direction of a different speaker altogether. Thus, caution should be exercised when applying this tool. |
2310.14129 | Pan Xu | Tianyuan Jin, Yu Yang, Jing Tang, Xiaokui Xiao, Pan Xu | Optimal Batched Best Arm Identification | 32 pages, 1 figure, 3 tables | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study the batched best arm identification (BBAI) problem, where the
learner's goal is to identify the best arm while switching the policy as less
as possible. In particular, we aim to find the best arm with probability
$1-\delta$ for some small constant $\delta>0$ while minimizing both the sample
complexity (total number of arm pulls) and the batch complexity (total number
of batches). We propose the three-batch best arm identification (Tri-BBAI)
algorithm, which is the first batched algorithm that achieves the optimal
sample complexity in the asymptotic setting (i.e., $\delta\rightarrow 0$) and
runs only in at most $3$ batches. Based on Tri-BBAI, we further propose the
almost optimal batched best arm identification (Opt-BBAI) algorithm, which is
the first algorithm that achieves the near-optimal sample and batch complexity
in the non-asymptotic setting (i.e., $\delta>0$ is arbitrarily fixed), while
enjoying the same batch and sample complexity as Tri-BBAI when $\delta$ tends
to zero. Moreover, in the non-asymptotic setting, the complexity of previous
batch algorithms is usually conditioned on the event that the best arm is
returned (with a probability of at least $1-\delta$), which is potentially
unbounded in cases where a sub-optimal arm is returned. In contrast, the
complexity of Opt-BBAI does not rely on such an event. This is achieved through
a novel procedure that we design for checking whether the best arm is
eliminated, which is of independent interest.
| [
{
"created": "Sat, 21 Oct 2023 22:55:50 GMT",
"version": "v1"
}
] | 2023-10-24 | [
[
"Jin",
"Tianyuan",
""
],
[
"Yang",
"Yu",
""
],
[
"Tang",
"Jing",
""
],
[
"Xiao",
"Xiaokui",
""
],
[
"Xu",
"Pan",
""
]
] | We study the batched best arm identification (BBAI) problem, where the learner's goal is to identify the best arm while switching the policy as less as possible. In particular, we aim to find the best arm with probability $1-\delta$ for some small constant $\delta>0$ while minimizing both the sample complexity (total number of arm pulls) and the batch complexity (total number of batches). We propose the three-batch best arm identification (Tri-BBAI) algorithm, which is the first batched algorithm that achieves the optimal sample complexity in the asymptotic setting (i.e., $\delta\rightarrow 0$) and runs only in at most $3$ batches. Based on Tri-BBAI, we further propose the almost optimal batched best arm identification (Opt-BBAI) algorithm, which is the first algorithm that achieves the near-optimal sample and batch complexity in the non-asymptotic setting (i.e., $\delta>0$ is arbitrarily fixed), while enjoying the same batch and sample complexity as Tri-BBAI when $\delta$ tends to zero. Moreover, in the non-asymptotic setting, the complexity of previous batch algorithms is usually conditioned on the event that the best arm is returned (with a probability of at least $1-\delta$), which is potentially unbounded in cases where a sub-optimal arm is returned. In contrast, the complexity of Opt-BBAI does not rely on such an event. This is achieved through a novel procedure that we design for checking whether the best arm is eliminated, which is of independent interest. |
2407.12170 | Sean MacAvaney | Xuejun Chang, Debabrata Mishra, Craig Macdonald, Sean MacAvaney | Neural Passage Quality Estimation for Static Pruning | SIGIR 2024 | null | 10.1145/3626772.3657765 | null | cs.IR | http://creativecommons.org/licenses/by/4.0/ | Neural networks -- especially those that use large, pre-trained language
models -- have improved search engines in various ways. Most prominently, they
can estimate the relevance of a passage or document to a user's query. In this
work, we depart from this direction by exploring whether neural networks can
effectively predict which of a document's passages are unlikely to be relevant
to any query submitted to the search engine. We refer to this query-agnostic
estimation of passage relevance as a passage's quality. We find that our novel
methods for estimating passage quality allow passage corpora to be pruned
considerably while maintaining statistically equivalent effectiveness; our best
methods can consistently prune >25% of passages in a corpora, across various
retrieval pipelines. Such substantial pruning reduces the operating costs of
neural search engines in terms of computing resources, power usage, and carbon
footprint -- both when processing queries (thanks to a smaller index size) and
when indexing (lightweight models can prune low-quality passages prior to the
costly dense or learned sparse encoding step). This work sets the stage for
developing more advanced neural "learning-what-to-index" methods.
| [
{
"created": "Tue, 16 Jul 2024 20:47:54 GMT",
"version": "v1"
}
] | 2024-07-18 | [
[
"Chang",
"Xuejun",
""
],
[
"Mishra",
"Debabrata",
""
],
[
"Macdonald",
"Craig",
""
],
[
"MacAvaney",
"Sean",
""
]
] | Neural networks -- especially those that use large, pre-trained language models -- have improved search engines in various ways. Most prominently, they can estimate the relevance of a passage or document to a user's query. In this work, we depart from this direction by exploring whether neural networks can effectively predict which of a document's passages are unlikely to be relevant to any query submitted to the search engine. We refer to this query-agnostic estimation of passage relevance as a passage's quality. We find that our novel methods for estimating passage quality allow passage corpora to be pruned considerably while maintaining statistically equivalent effectiveness; our best methods can consistently prune >25% of passages in a corpora, across various retrieval pipelines. Such substantial pruning reduces the operating costs of neural search engines in terms of computing resources, power usage, and carbon footprint -- both when processing queries (thanks to a smaller index size) and when indexing (lightweight models can prune low-quality passages prior to the costly dense or learned sparse encoding step). This work sets the stage for developing more advanced neural "learning-what-to-index" methods. |
2205.15906 | Malsha V Perera | Malsha V. Perera, Wele Gedara Chaminda Bandara, Jeya Maria Jose
Valanarasu, and Vishal M. Patel | SAR Despeckling Using Overcomplete Convolutional Networks | Accepted to International Geoscience and Remote Sensing Symposium
(IGARSS), 2022. Our code is available at
https://github.com/malshaV/sar_overcomplete | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Synthetic Aperture Radar (SAR) despeckling is an important problem in remote
sensing as speckle degrades SAR images, affecting downstream tasks like
detection and segmentation. Recent studies show that convolutional neural
networks(CNNs) outperform classical despeckling methods. Traditional CNNs try
to increase the receptive field size as the network goes deeper, thus
extracting global features. However,speckle is relatively small, and increasing
receptive field does not help in extracting speckle features. This study
employs an overcomplete CNN architecture to focus on learning low-level
features by restricting the receptive field. The proposed network consists of
an overcomplete branch to focus on the local structures and an undercomplete
branch that focuses on the global structures. We show that the proposed network
improves despeckling performance compared to recent despeckling methods on
synthetic and real SAR images.
| [
{
"created": "Tue, 31 May 2022 15:55:37 GMT",
"version": "v1"
}
] | 2022-06-01 | [
[
"Perera",
"Malsha V.",
""
],
[
"Bandara",
"Wele Gedara Chaminda",
""
],
[
"Valanarasu",
"Jeya Maria Jose",
""
],
[
"Patel",
"Vishal M.",
""
]
] | Synthetic Aperture Radar (SAR) despeckling is an important problem in remote sensing as speckle degrades SAR images, affecting downstream tasks like detection and segmentation. Recent studies show that convolutional neural networks(CNNs) outperform classical despeckling methods. Traditional CNNs try to increase the receptive field size as the network goes deeper, thus extracting global features. However,speckle is relatively small, and increasing receptive field does not help in extracting speckle features. This study employs an overcomplete CNN architecture to focus on learning low-level features by restricting the receptive field. The proposed network consists of an overcomplete branch to focus on the local structures and an undercomplete branch that focuses on the global structures. We show that the proposed network improves despeckling performance compared to recent despeckling methods on synthetic and real SAR images. |
1805.00443 | Lucie Jacquin | Pierre-Michel Riccio (LGI2P) | Une approche pour mieux appr{\'e}hender l'alt{\'e}rit{\'e} en SIC | in French | Colloque Communication, Organisation, Soci{\'e}t{\'e} du Savoir et
Information (COSSI) 15-17 juin 2016 , Jun 2016, Montpellier, France | null | null | cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a novel approach that aims: to facilitate the
building of teams relying on notions of skills, of motivation or of potential;
identify the requested upgrade effort to improve his own knowledge and join a
team; or create more suitable devices for a target population. The whole forms
a toolbox that seems appropriate to facilitate better recognition of otherness.
| [
{
"created": "Fri, 20 Apr 2018 11:20:51 GMT",
"version": "v1"
}
] | 2018-05-02 | [
[
"Riccio",
"Pierre-Michel",
"",
"LGI2P"
]
] | In this paper, we propose a novel approach that aims: to facilitate the building of teams relying on notions of skills, of motivation or of potential; identify the requested upgrade effort to improve his own knowledge and join a team; or create more suitable devices for a target population. The whole forms a toolbox that seems appropriate to facilitate better recognition of otherness. |
2401.14079 | Tobias Eisenreich | Tobias Eisenreich, Sandro Speth, Stefan Wagner | From Requirements to Architecture: An AI-Based Journey to
Semi-Automatically Generate Software Architectures | 4 pages, vision paper, submitted to the ICSE workshop Designing2024 | null | 10.1145/3643660.3643942 | null | cs.SE cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Designing domain models and software architectures represents a significant
challenge in software development, as the resulting architectures play a vital
role in fulfilling the system's quality of service. Due to time pressure,
architects often model only one architecture based on their known limited
domain understanding, patterns, and experience instead of thoroughly analyzing
the domain and evaluating multiple candidates, selecting the best fitting.
Existing approaches try to generate domain models based on requirements, but
still require time-consuming manual effort to achieve good results. Therefore,
in this vision paper, we propose a method to generate software architecture
candidates semi-automatically based on requirements using artificial
intelligence techniques. We further envision an automatic evaluation and
trade-off analysis of the generated architecture candidates using, e.g., the
architecture trade-off analysis method combined with large language models and
quantitative analyses. To evaluate this approach, we aim to analyze the quality
of the generated architecture models and the efficiency and effectiveness of
our proposed process by conducting qualitative studies.
| [
{
"created": "Thu, 25 Jan 2024 10:56:58 GMT",
"version": "v1"
}
] | 2024-02-02 | [
[
"Eisenreich",
"Tobias",
""
],
[
"Speth",
"Sandro",
""
],
[
"Wagner",
"Stefan",
""
]
] | Designing domain models and software architectures represents a significant challenge in software development, as the resulting architectures play a vital role in fulfilling the system's quality of service. Due to time pressure, architects often model only one architecture based on their known limited domain understanding, patterns, and experience instead of thoroughly analyzing the domain and evaluating multiple candidates, selecting the best fitting. Existing approaches try to generate domain models based on requirements, but still require time-consuming manual effort to achieve good results. Therefore, in this vision paper, we propose a method to generate software architecture candidates semi-automatically based on requirements using artificial intelligence techniques. We further envision an automatic evaluation and trade-off analysis of the generated architecture candidates using, e.g., the architecture trade-off analysis method combined with large language models and quantitative analyses. To evaluate this approach, we aim to analyze the quality of the generated architecture models and the efficiency and effectiveness of our proposed process by conducting qualitative studies. |
1406.6114 | Sakthithasan Sripirakas | Sakthithasan Sripirakas and Russel Pears | Mining Recurrent Concepts in Data Streams using the Discrete Fourier
Transform | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this research we address the problem of capturing recurring concepts in a
data stream environment. Recurrence capture enables the re-use of previously
learned classifiers without the need for re-learning while providing for better
accuracy during the concept recurrence interval. We capture concepts by
applying the Discrete Fourier Transform (DFT) to Decision Tree classifiers to
obtain highly compressed versions of the trees at concept drift points in the
stream and store such trees in a repository for future use. Our empirical
results on real world and synthetic data exhibiting varying degrees of
recurrence show that the Fourier compressed trees are more robust to noise and
are able to capture recurring concepts with higher precision than a meta
learning approach that chooses to re-use classifiers in their originally
occurring form.
| [
{
"created": "Tue, 24 Jun 2014 00:48:23 GMT",
"version": "v1"
}
] | 2014-06-25 | [
[
"Sripirakas",
"Sakthithasan",
""
],
[
"Pears",
"Russel",
""
]
] | In this research we address the problem of capturing recurring concepts in a data stream environment. Recurrence capture enables the re-use of previously learned classifiers without the need for re-learning while providing for better accuracy during the concept recurrence interval. We capture concepts by applying the Discrete Fourier Transform (DFT) to Decision Tree classifiers to obtain highly compressed versions of the trees at concept drift points in the stream and store such trees in a repository for future use. Our empirical results on real world and synthetic data exhibiting varying degrees of recurrence show that the Fourier compressed trees are more robust to noise and are able to capture recurring concepts with higher precision than a meta learning approach that chooses to re-use classifiers in their originally occurring form. |
2402.03396 | Yifeng He | Yifeng He, Jiabo Huang, Yuyang Rong, Yiwen Guo, Ethan Wang, Hao Chen | UniTSyn: A Large-Scale Dataset Capable of Enhancing the Prowess of Large
Language Models for Program Testing | 8 pages, 5 figures | null | null | null | cs.SE cs.AI cs.CL cs.CR cs.LG | http://creativecommons.org/licenses/by/4.0/ | The remarkable capability of large language models (LLMs) in generating
high-quality code has drawn increasing attention in the software testing
community. However, existing code LLMs often demonstrate unsatisfactory
capabilities in generating accurate and complete tests since they were trained
on code snippets collected without differentiating between code for testing
purposes and other code. In this paper, we present a large-scale dataset
UniTSyn, which is capable of enhancing the prowess of LLMs for Unit Test
Synthesis. Associating tests with the tested functions is crucial for LLMs to
infer the expected behavior and the logic paths to be verified. By leveraging
Language Server Protocol, UniTSyn achieves the challenging goal of collecting
focal-test pairs without per-project execution setups or per-language
heuristics that tend to be fragile and difficult to scale. It contains 2.7
million focal-test pairs across five mainstream programming languages, making
it possible to be utilized for enhancing the test generation ability of LLMs.
The details of UniTSyn can be found in Table 1. Our experiments demonstrate
that, by building an autoregressive model based on UniTSyn, we can achieve
significant benefits in learning and understanding unit test representations,
resulting in improved generation accuracy and code coverage across all
evaluated programming languages. Code and data will be publicly available.
| [
{
"created": "Sun, 4 Feb 2024 22:48:05 GMT",
"version": "v1"
}
] | 2024-02-07 | [
[
"He",
"Yifeng",
""
],
[
"Huang",
"Jiabo",
""
],
[
"Rong",
"Yuyang",
""
],
[
"Guo",
"Yiwen",
""
],
[
"Wang",
"Ethan",
""
],
[
"Chen",
"Hao",
""
]
] | The remarkable capability of large language models (LLMs) in generating high-quality code has drawn increasing attention in the software testing community. However, existing code LLMs often demonstrate unsatisfactory capabilities in generating accurate and complete tests since they were trained on code snippets collected without differentiating between code for testing purposes and other code. In this paper, we present a large-scale dataset UniTSyn, which is capable of enhancing the prowess of LLMs for Unit Test Synthesis. Associating tests with the tested functions is crucial for LLMs to infer the expected behavior and the logic paths to be verified. By leveraging Language Server Protocol, UniTSyn achieves the challenging goal of collecting focal-test pairs without per-project execution setups or per-language heuristics that tend to be fragile and difficult to scale. It contains 2.7 million focal-test pairs across five mainstream programming languages, making it possible to be utilized for enhancing the test generation ability of LLMs. The details of UniTSyn can be found in Table 1. Our experiments demonstrate that, by building an autoregressive model based on UniTSyn, we can achieve significant benefits in learning and understanding unit test representations, resulting in improved generation accuracy and code coverage across all evaluated programming languages. Code and data will be publicly available. |
2008.04701 | Matthijs Maas | John-Clark Levin and Matthijs M. Maas | Roadmap to a Roadmap: How Could We Tell When AGI is a 'Manhattan
Project' Away? | null | null | null | null | cs.CY | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper argues that at a certain point in research toward AGI, the problem
may become well-enough theorized that a clear roadmap exists for achieving it,
such that a Manhattan Project-like effort could greatly shorten the time to
completion. If state actors perceive that this threshold has been crossed,
their incentives around openness and international cooperation may shift rather
suddenly, with serious implications for AI risks and the stability of
international AI governance regimes. The paper characterizes how such a
'runway' period would be qualitatively different from preceding stages of AI
research, and accordingly proposes a research program aimed at assessing how
close the field of AI is to such a threshold - that is, it calls for the
formulation of a 'roadmap to the roadmap.'
| [
{
"created": "Thu, 6 Aug 2020 06:07:47 GMT",
"version": "v1"
}
] | 2020-08-12 | [
[
"Levin",
"John-Clark",
""
],
[
"Maas",
"Matthijs M.",
""
]
] | This paper argues that at a certain point in research toward AGI, the problem may become well-enough theorized that a clear roadmap exists for achieving it, such that a Manhattan Project-like effort could greatly shorten the time to completion. If state actors perceive that this threshold has been crossed, their incentives around openness and international cooperation may shift rather suddenly, with serious implications for AI risks and the stability of international AI governance regimes. The paper characterizes how such a 'runway' period would be qualitatively different from preceding stages of AI research, and accordingly proposes a research program aimed at assessing how close the field of AI is to such a threshold - that is, it calls for the formulation of a 'roadmap to the roadmap.' |
cs/0701134 | Wenbing Zhao | Wenbing Zhao | Byzantine Fault Tolerance for Nondeterministic Applications | To appear in the proceedings of the 3rd IEEE International Symposium
on Dependable, Autonomic and Secure Computing, 2007 | null | 10.1109/DASC.2007.11 | null | cs.DC | null | All practical applications contain some degree of nondeterminism. When such
applications are replicated to achieve Byzantine fault tolerance (BFT), their
nondeterministic operations must be controlled to ensure replica consistency.
To the best of our knowledge, only the most simplistic types of replica
nondeterminism have been dealt with. Furthermore, there lacks a systematic
approach to handling common types of nondeterminism. In this paper, we propose
a classification of common types of replica nondeterminism with respect to the
requirement of achieving Byzantine fault tolerance, and describe the design and
implementation of the core mechanisms necessary to handle such nondeterminism
within a Byzantine fault tolerance framework.
| [
{
"created": "Sun, 21 Jan 2007 20:44:52 GMT",
"version": "v1"
},
{
"created": "Wed, 1 Aug 2007 04:51:53 GMT",
"version": "v2"
}
] | 2016-11-15 | [
[
"Zhao",
"Wenbing",
""
]
] | All practical applications contain some degree of nondeterminism. When such applications are replicated to achieve Byzantine fault tolerance (BFT), their nondeterministic operations must be controlled to ensure replica consistency. To the best of our knowledge, only the most simplistic types of replica nondeterminism have been dealt with. Furthermore, there lacks a systematic approach to handling common types of nondeterminism. In this paper, we propose a classification of common types of replica nondeterminism with respect to the requirement of achieving Byzantine fault tolerance, and describe the design and implementation of the core mechanisms necessary to handle such nondeterminism within a Byzantine fault tolerance framework. |
1801.03911 | Sahil Garg | Sahil Garg and Greg Ver Steeg and Aram Galstyan | Stochastic Learning of Nonstationary Kernels for Natural Language
Modeling | null | null | null | null | cs.CL cs.IR cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Natural language processing often involves computations with semantic or
syntactic graphs to facilitate sophisticated reasoning based on structural
relationships. While convolution kernels provide a powerful tool for comparing
graph structure based on node (word) level relationships, they are difficult to
customize and can be computationally expensive. We propose a generalization of
convolution kernels, with a nonstationary model, for better expressibility of
natural languages in supervised settings. For a scalable learning of the
parameters introduced with our model, we propose a novel algorithm that
leverages stochastic sampling on k-nearest neighbor graphs, along with
approximations based on locality-sensitive hashing. We demonstrate the
advantages of our approach on a challenging real-world (structured inference)
problem of automatically extracting biological models from the text of
scientific papers.
| [
{
"created": "Thu, 11 Jan 2018 18:24:02 GMT",
"version": "v1"
},
{
"created": "Thu, 1 Feb 2018 21:41:27 GMT",
"version": "v2"
}
] | 2018-02-13 | [
[
"Garg",
"Sahil",
""
],
[
"Steeg",
"Greg Ver",
""
],
[
"Galstyan",
"Aram",
""
]
] | Natural language processing often involves computations with semantic or syntactic graphs to facilitate sophisticated reasoning based on structural relationships. While convolution kernels provide a powerful tool for comparing graph structure based on node (word) level relationships, they are difficult to customize and can be computationally expensive. We propose a generalization of convolution kernels, with a nonstationary model, for better expressibility of natural languages in supervised settings. For a scalable learning of the parameters introduced with our model, we propose a novel algorithm that leverages stochastic sampling on k-nearest neighbor graphs, along with approximations based on locality-sensitive hashing. We demonstrate the advantages of our approach on a challenging real-world (structured inference) problem of automatically extracting biological models from the text of scientific papers. |
1912.12430 | Theofilos Triommatis | Theofilos Triommatis and Aris Pagourtzis | Approximate #Knapsack Computations to Count Semi-Fair Allocations | null | null | null | null | cs.CC | http://creativecommons.org/licenses/by-nc-sa/4.0/ | In this paper, we study the problem of counting the number of different
knapsack solutions with a prescribed cardinality. We present an FPTAS for this
problem, based on dynamic programming. We also introduce two different types of
semi-fair allocations of indivisible goods between two players. By semi-fair
allocations, we mean allocations that ensure that at least one of the two
players will be free of envy. We study the problem of counting such allocations
and we provide FPTASs for both types, by employing our FPTAS for the prescribed
cardinality knapsack problem.
| [
{
"created": "Sat, 28 Dec 2019 09:48:32 GMT",
"version": "v1"
}
] | 2020-01-01 | [
[
"Triommatis",
"Theofilos",
""
],
[
"Pagourtzis",
"Aris",
""
]
] | In this paper, we study the problem of counting the number of different knapsack solutions with a prescribed cardinality. We present an FPTAS for this problem, based on dynamic programming. We also introduce two different types of semi-fair allocations of indivisible goods between two players. By semi-fair allocations, we mean allocations that ensure that at least one of the two players will be free of envy. We study the problem of counting such allocations and we provide FPTASs for both types, by employing our FPTAS for the prescribed cardinality knapsack problem. |
2406.04823 | David Samuel | David Samuel | BERTs are Generative In-Context Learners | 21 pages, preprint | null | null | null | cs.CL cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper explores the in-context learning capabilities of masked language
models, challenging the common view that this ability does not 'emerge' in
them. We present an embarrassingly simple inference technique that enables
DeBERTa to operate as a generative model without any additional training. Our
findings demonstrate that DeBERTa can match and even surpass GPT-3, its
contemporary that famously introduced the paradigm of in-context learning. The
comparative analysis reveals that the masked and causal language models behave
very differently, as they clearly outperform each other on different categories
of tasks. This suggests that there is great potential for a hybrid training
approach that takes advantage of the strengths of both training objectives.
| [
{
"created": "Fri, 7 Jun 2024 10:48:45 GMT",
"version": "v1"
}
] | 2024-06-10 | [
[
"Samuel",
"David",
""
]
] | This paper explores the in-context learning capabilities of masked language models, challenging the common view that this ability does not 'emerge' in them. We present an embarrassingly simple inference technique that enables DeBERTa to operate as a generative model without any additional training. Our findings demonstrate that DeBERTa can match and even surpass GPT-3, its contemporary that famously introduced the paradigm of in-context learning. The comparative analysis reveals that the masked and causal language models behave very differently, as they clearly outperform each other on different categories of tasks. This suggests that there is great potential for a hybrid training approach that takes advantage of the strengths of both training objectives. |
cs/0511063 | Michele Finelli | Michele Finelli | Pathwords: a user-friendly schema for common passwords management | null | null | null | null | cs.CR | null | Many computer-based authentication schemata are based on pass- words. Logging
on a computer, reading email, accessing content on a web server are all
examples of applications where the identification of the user is usually
accomplished matching the data provided by the user with data known by the
application.
Such a widespread approach relies on some assumptions, whose satisfaction is
of foremost importance to guarantee the robustness of the solution. Some of
these assumptions, like having a "secure" chan- nel to transmit data, or having
sound algorithms to check the correct- ness of the data, are not addressed by
this paper. We will focus on two simple issues: the problem of using adequate
passwords and the problem of managing passwords.
The proposed solution, the pathword, is a method that guarantees:
1 that the passwords generated with the help of a pathword are adequate (i.e.
that they are not easy to guess),
2 that managing pathwords is more user friendly than managing passwords and
that pathwords are less amenable to problems typical of passwords.
| [
{
"created": "Wed, 16 Nov 2005 17:46:48 GMT",
"version": "v1"
}
] | 2007-05-23 | [
[
"Finelli",
"Michele",
""
]
] | Many computer-based authentication schemata are based on pass- words. Logging on a computer, reading email, accessing content on a web server are all examples of applications where the identification of the user is usually accomplished matching the data provided by the user with data known by the application. Such a widespread approach relies on some assumptions, whose satisfaction is of foremost importance to guarantee the robustness of the solution. Some of these assumptions, like having a "secure" chan- nel to transmit data, or having sound algorithms to check the correct- ness of the data, are not addressed by this paper. We will focus on two simple issues: the problem of using adequate passwords and the problem of managing passwords. The proposed solution, the pathword, is a method that guarantees: 1 that the passwords generated with the help of a pathword are adequate (i.e. that they are not easy to guess), 2 that managing pathwords is more user friendly than managing passwords and that pathwords are less amenable to problems typical of passwords. |
2402.01386 | Zeeshan Rasheed Mr | Zeeshan Rasheed, Muhammad Waseem, Aakash Ahmad, Kai-Kristian Kemell,
Wang Xiaofeng, Anh Nguyen Duc, Pekka Abrahamsson | Can Large Language Models Serve as Data Analysts? A Multi-Agent Assisted
Approach for Qualitative Data Analysis | 9 pages and 2 figures | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recent advancements in Large Language Models (LLMs) have enabled
collaborative human-bot interactions in Software Engineering (SE), similar to
many other professions. However, the potential benefits and implications of
incorporating LLMs into qualitative data analysis in SE have not been
completely explored. For instance, conducting qualitative data analysis
manually can be a time-consuming, effort-intensive, and error-prone task for
researchers. LLM-based solutions, such as generative AI models trained on
massive datasets, can be utilized to automate tasks in software development as
well as in qualitative data analysis. To this end, we utilized LLMs to automate
and expedite the qualitative data analysis processes. We employed a multi-agent
model, where each agent was tasked with executing distinct, individual research
related activities. Our proposed model interpreted large quantities of textual
documents and interview transcripts to perform several common tasks used in
qualitative analysis. The results show that this technical assistant speeds up
significantly the data analysis process, enabling researchers to manage larger
datasets much more effectively. Furthermore, this approach introduces a new
dimension of scalability and accuracy in qualitative research, potentially
transforming data interpretation methodologies in SE.
| [
{
"created": "Fri, 2 Feb 2024 13:10:46 GMT",
"version": "v1"
}
] | 2024-02-05 | [
[
"Rasheed",
"Zeeshan",
""
],
[
"Waseem",
"Muhammad",
""
],
[
"Ahmad",
"Aakash",
""
],
[
"Kemell",
"Kai-Kristian",
""
],
[
"Xiaofeng",
"Wang",
""
],
[
"Duc",
"Anh Nguyen",
""
],
[
"Abrahamsson",
"Pekka",
""
]
] | Recent advancements in Large Language Models (LLMs) have enabled collaborative human-bot interactions in Software Engineering (SE), similar to many other professions. However, the potential benefits and implications of incorporating LLMs into qualitative data analysis in SE have not been completely explored. For instance, conducting qualitative data analysis manually can be a time-consuming, effort-intensive, and error-prone task for researchers. LLM-based solutions, such as generative AI models trained on massive datasets, can be utilized to automate tasks in software development as well as in qualitative data analysis. To this end, we utilized LLMs to automate and expedite the qualitative data analysis processes. We employed a multi-agent model, where each agent was tasked with executing distinct, individual research related activities. Our proposed model interpreted large quantities of textual documents and interview transcripts to perform several common tasks used in qualitative analysis. The results show that this technical assistant speeds up significantly the data analysis process, enabling researchers to manage larger datasets much more effectively. Furthermore, this approach introduces a new dimension of scalability and accuracy in qualitative research, potentially transforming data interpretation methodologies in SE. |
1703.08206 | Manuel Peuster | Manuel Peuster and Holger Karl | Understand Your Chains: Towards Performance Profile-based Network
Service Management | Submitted to and accepted by the European Workshop on Software
Defined Networks (EWSDN) 2016 | null | null | null | cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Allocating resources to virtualized network functions and services to meet
service level agreements is a challenging task for NFV management and
orchestration systems. This becomes even more challenging when agile
development methodologies, like DevOps, are applied. In such scenarios,
management and orchestration systems are continuously facing new versions of
functions and services which makes it hard to decide how much resources have to
be allocated to them to provide the expected service performance. One solution
for this problem is to support resource allocation decisions with performance
behavior information obtained by profiling techniques applied to such network
functions and services.
In this position paper, we analyze and discuss the components needed to
generate such performance behavior information within the NFV DevOps workflow.
We also outline research questions that identify open issues and missing pieces
for a fully integrated NFV profiling solution. Further, we introduce a novel
profiling mechanism that is able to profile virtualized network functions and
entire network service chains under different resource constraints before they
are deployed on production infrastructure.
| [
{
"created": "Thu, 23 Mar 2017 19:12:09 GMT",
"version": "v1"
}
] | 2017-03-27 | [
[
"Peuster",
"Manuel",
""
],
[
"Karl",
"Holger",
""
]
] | Allocating resources to virtualized network functions and services to meet service level agreements is a challenging task for NFV management and orchestration systems. This becomes even more challenging when agile development methodologies, like DevOps, are applied. In such scenarios, management and orchestration systems are continuously facing new versions of functions and services which makes it hard to decide how much resources have to be allocated to them to provide the expected service performance. One solution for this problem is to support resource allocation decisions with performance behavior information obtained by profiling techniques applied to such network functions and services. In this position paper, we analyze and discuss the components needed to generate such performance behavior information within the NFV DevOps workflow. We also outline research questions that identify open issues and missing pieces for a fully integrated NFV profiling solution. Further, we introduce a novel profiling mechanism that is able to profile virtualized network functions and entire network service chains under different resource constraints before they are deployed on production infrastructure. |
2403.19449 | Marcin Hoffmann | Marcin Hoffmann, Pawe{\l} Kryszkiewicz | O-RAN for Energy-Efficient Serving Cluster Formulation in User-Centric
Cell-Free MMIMO | Accepted for presentation during The 2nd Workshop on Next-generation
Open and Programmable Radio Access Networks (NG-OPERA), organized in
conjunction with IEEE International Conference on Computer Communications,
May 20, 2024 | null | null | null | cs.IT cs.NI math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The 6G Massive Multiple-Input Multiple-Output (MMIMO) networks can follow the
so-called User-Centric Cell-Free (UCCF) architecture, where a single user is
served by multiple Access Points (APs) coordinated by the Central Processing
Unit (CPU). In this paper, we propose how O-RAN functionalities, i.e.,
rApp-xApp pair, can be used for energy-efficient Serving Cluster Formulation
(SCF). Simulation studies show up to 37\% gain in Energy Efficiency (EE) of the
proposed solution over the state-of-the-art Network-Centric (NC) designs.
| [
{
"created": "Thu, 28 Mar 2024 14:17:19 GMT",
"version": "v1"
}
] | 2024-03-29 | [
[
"Hoffmann",
"Marcin",
""
],
[
"Kryszkiewicz",
"Paweł",
""
]
] | The 6G Massive Multiple-Input Multiple-Output (MMIMO) networks can follow the so-called User-Centric Cell-Free (UCCF) architecture, where a single user is served by multiple Access Points (APs) coordinated by the Central Processing Unit (CPU). In this paper, we propose how O-RAN functionalities, i.e., rApp-xApp pair, can be used for energy-efficient Serving Cluster Formulation (SCF). Simulation studies show up to 37\% gain in Energy Efficiency (EE) of the proposed solution over the state-of-the-art Network-Centric (NC) designs. |
2301.06923 | Zhibo Zhang | Zhibo Zhang, Sani Umar, Ahmed Y. Al Hammadi, Sangyoung Yoon, Ernesto
Damiani, Claudio Agostino Ardagna, Nicola Bena, and Chan Yeob Yeun | Explainable Data Poison Attacks on Human Emotion Evaluation Systems
based on EEG Signals | null | IEEE Access 2023 | 10.1109/ACCESS.2023.3245813 | null | cs.LG eess.SP | http://creativecommons.org/licenses/by/4.0/ | The major aim of this paper is to explain the data poisoning attacks using
label-flipping during the training stage of the electroencephalogram (EEG)
signal-based human emotion evaluation systems deploying Machine Learning models
from the attackers' perspective. Human emotion evaluation using EEG signals has
consistently attracted a lot of research attention. The identification of human
emotional states based on EEG signals is effective to detect potential internal
threats caused by insider individuals. Nevertheless, EEG signal-based human
emotion evaluation systems have shown several vulnerabilities to data poison
attacks. The findings of the experiments demonstrate that the suggested data
poison assaults are model-independently successful, although various models
exhibit varying levels of resilience to the attacks. In addition, the data
poison attacks on the EEG signal-based human emotion evaluation systems are
explained with several Explainable Artificial Intelligence (XAI) methods,
including Shapley Additive Explanation (SHAP) values, Local Interpretable
Model-agnostic Explanations (LIME), and Generated Decision Trees. And the codes
of this paper are publicly available on GitHub.
| [
{
"created": "Tue, 17 Jan 2023 14:44:46 GMT",
"version": "v1"
}
] | 2023-03-15 | [
[
"Zhang",
"Zhibo",
""
],
[
"Umar",
"Sani",
""
],
[
"Hammadi",
"Ahmed Y. Al",
""
],
[
"Yoon",
"Sangyoung",
""
],
[
"Damiani",
"Ernesto",
""
],
[
"Ardagna",
"Claudio Agostino",
""
],
[
"Bena",
"Nicola",
""
],
[
"Yeun",
"Chan Yeob",
""
]
] | The major aim of this paper is to explain the data poisoning attacks using label-flipping during the training stage of the electroencephalogram (EEG) signal-based human emotion evaluation systems deploying Machine Learning models from the attackers' perspective. Human emotion evaluation using EEG signals has consistently attracted a lot of research attention. The identification of human emotional states based on EEG signals is effective to detect potential internal threats caused by insider individuals. Nevertheless, EEG signal-based human emotion evaluation systems have shown several vulnerabilities to data poison attacks. The findings of the experiments demonstrate that the suggested data poison assaults are model-independently successful, although various models exhibit varying levels of resilience to the attacks. In addition, the data poison attacks on the EEG signal-based human emotion evaluation systems are explained with several Explainable Artificial Intelligence (XAI) methods, including Shapley Additive Explanation (SHAP) values, Local Interpretable Model-agnostic Explanations (LIME), and Generated Decision Trees. And the codes of this paper are publicly available on GitHub. |
2005.08625 | Na Li | Na Li, Xinbo Zhao, Chong Ma | JointsGait:A model-based Gait Recognition Method based on Gait Graph
Convolutional Networks and Joints Relationship Pyramid Mapping | The paper format was changed and experiments on other databases were
added. The format and page layout were changed greatly | null | null | null | cs.CV eess.IV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Gait, as one of unique biometric features, has the advantage of being
recognized from a long distance away, can be widely used in public security.
Considering 3D pose estimation is more challenging than 2D pose estimation in
practice , we research on using 2D joints to recognize gait in this paper, and
a new model-based gait recognition method JointsGait is put forward to extract
gait information from 2D human body joints. Appearance-based gait recognition
algorithms are prevalent before. However, appearance features suffer from
external factors which can cause drastic appearance variations, e.g. clothing.
Unlike previous approaches, JointsGait firstly extracted spatio-temporal
features from 2D joints using gait graph convolutional networks, which are less
interfered by external factors. Secondly, Joints Relationship Pyramid Mapping
(JRPM) are proposed to map spatio-temporal gait features into a discriminative
feature space with biological advantages according to the relationship of human
joints when people are walking at various scales. Finally, we design a fusion
loss strategy to help the joints features to be insensitive to cross-view. Our
method is evaluated on two large datasets, Kinect Gait Biometry Dataset and
CASIA-B. On Kinect Gait Biometry Dataset database, JointsGait only uses
corresponding 2D coordinates of joints, but achieves satisfactory recognition
accuracy compared with those model-based algorithms using 3D joints. On CASIA-B
database, the proposed method greatly outperforms advanced model-based methods
in all walking conditions, even performs superior to state-of-art
appearance-based methods when clothing seriously affect people's appearance.
The experimental results demonstrate that JointsGait achieves the state-of-art
performance despite the low dimensional feature (2D body joints) and is less
affected by the view variations and clothing variation.
| [
{
"created": "Mon, 27 Apr 2020 08:30:37 GMT",
"version": "v1"
},
{
"created": "Wed, 9 Dec 2020 09:12:03 GMT",
"version": "v2"
}
] | 2020-12-10 | [
[
"Li",
"Na",
""
],
[
"Zhao",
"Xinbo",
""
],
[
"Ma",
"Chong",
""
]
] | Gait, as one of unique biometric features, has the advantage of being recognized from a long distance away, can be widely used in public security. Considering 3D pose estimation is more challenging than 2D pose estimation in practice , we research on using 2D joints to recognize gait in this paper, and a new model-based gait recognition method JointsGait is put forward to extract gait information from 2D human body joints. Appearance-based gait recognition algorithms are prevalent before. However, appearance features suffer from external factors which can cause drastic appearance variations, e.g. clothing. Unlike previous approaches, JointsGait firstly extracted spatio-temporal features from 2D joints using gait graph convolutional networks, which are less interfered by external factors. Secondly, Joints Relationship Pyramid Mapping (JRPM) are proposed to map spatio-temporal gait features into a discriminative feature space with biological advantages according to the relationship of human joints when people are walking at various scales. Finally, we design a fusion loss strategy to help the joints features to be insensitive to cross-view. Our method is evaluated on two large datasets, Kinect Gait Biometry Dataset and CASIA-B. On Kinect Gait Biometry Dataset database, JointsGait only uses corresponding 2D coordinates of joints, but achieves satisfactory recognition accuracy compared with those model-based algorithms using 3D joints. On CASIA-B database, the proposed method greatly outperforms advanced model-based methods in all walking conditions, even performs superior to state-of-art appearance-based methods when clothing seriously affect people's appearance. The experimental results demonstrate that JointsGait achieves the state-of-art performance despite the low dimensional feature (2D body joints) and is less affected by the view variations and clothing variation. |
2403.14679 | Lorenzo Pellegrini | Davide Maltoni, Lorenzo Pellegrini | Continual Learning by Three-Phase Consolidation | 13 pages, 2 figures, 8 tables. Preprint under review | null | null | null | cs.LG cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | TPC (Three-Phase Consolidation) is here introduced as a simple but effective
approach to continually learn new classes (and/or instances of known classes)
while controlling forgetting of previous knowledge. Each experience (a.k.a.
task) is learned in three phases characterized by different rules and learning
dynamics, aimed at removing the class-bias problem (due to class unbalancing)
and limiting gradient-based corrections to prevent forgetting of
underrepresented classes. Several experiments on complex datasets demonstrate
its accuracy and efficiency advantages over competitive existing approaches.
The algorithm and all the results presented in this paper are fully
reproducible thanks to its publication on the Avalanche open framework for
continual learning.
| [
{
"created": "Tue, 12 Mar 2024 15:31:14 GMT",
"version": "v1"
}
] | 2024-03-25 | [
[
"Maltoni",
"Davide",
""
],
[
"Pellegrini",
"Lorenzo",
""
]
] | TPC (Three-Phase Consolidation) is here introduced as a simple but effective approach to continually learn new classes (and/or instances of known classes) while controlling forgetting of previous knowledge. Each experience (a.k.a. task) is learned in three phases characterized by different rules and learning dynamics, aimed at removing the class-bias problem (due to class unbalancing) and limiting gradient-based corrections to prevent forgetting of underrepresented classes. Several experiments on complex datasets demonstrate its accuracy and efficiency advantages over competitive existing approaches. The algorithm and all the results presented in this paper are fully reproducible thanks to its publication on the Avalanche open framework for continual learning. |
1210.1357 | Bojin Zheng | Jun Qin, Hongrun Wu, Xiaonian Tong, Bojin Zheng | A quantitative method for determining the robustness of complex networks | null | Physica D 2013 253 85--90 | 10.1016/j.physd.2013.03.002 | null | cs.SI nlin.AO physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Most current studies estimate the invulnerability of complex networks using a
qualitative method that analyzes the inaccurate decay rate of network
efficiency. This method results in confusion over the invulnerability of
various types of complex networks. By normalizing network efficiency and
defining a baseline, this paper defines the invulnerability index as the
integral of the difference between the normalized network efficiency curve and
the baseline. This quantitative method seeks to establish a benchmark for the
robustness and fragility of networks and to measure network invulnerability
under both edge and node attacks. To validate the reliability of the proposed
method, three small-world networks were selected as test beds. The simulation
results indicate that the proposed invulnerability index can effectively and
accurately quantify network resilience. The index should provide a valuable
reference for determining network invulnerability in future research.
| [
{
"created": "Thu, 4 Oct 2012 09:46:26 GMT",
"version": "v1"
},
{
"created": "Sat, 23 Mar 2013 09:33:00 GMT",
"version": "v2"
}
] | 2014-02-18 | [
[
"Qin",
"Jun",
""
],
[
"Wu",
"Hongrun",
""
],
[
"Tong",
"Xiaonian",
""
],
[
"Zheng",
"Bojin",
""
]
] | Most current studies estimate the invulnerability of complex networks using a qualitative method that analyzes the inaccurate decay rate of network efficiency. This method results in confusion over the invulnerability of various types of complex networks. By normalizing network efficiency and defining a baseline, this paper defines the invulnerability index as the integral of the difference between the normalized network efficiency curve and the baseline. This quantitative method seeks to establish a benchmark for the robustness and fragility of networks and to measure network invulnerability under both edge and node attacks. To validate the reliability of the proposed method, three small-world networks were selected as test beds. The simulation results indicate that the proposed invulnerability index can effectively and accurately quantify network resilience. The index should provide a valuable reference for determining network invulnerability in future research. |
2107.13180 | Javier Naranjo-Alcazar | Javier Naranjo-Alcazar, Sergi Perez-Castanos, Aaron Lopez-Garcia,
Pedro Zuccarello, Maximo Cobos, Francesc J. Ferri | Squeeze-Excitation Convolutional Recurrent Neural Networks for
Audio-Visual Scene Classification | null | null | null | null | cs.MM cs.CV cs.SD eess.AS eess.IV | http://creativecommons.org/publicdomain/zero/1.0/ | The use of multiple and semantically correlated sources can provide
complementary information to each other that may not be evident when working
with individual modalities on their own. In this context, multi-modal models
can help producing more accurate and robust predictions in machine learning
tasks where audio-visual data is available. This paper presents a multi-modal
model for automatic scene classification that exploits simultaneously auditory
and visual information. The proposed approach makes use of two separate
networks which are respectively trained in isolation on audio and visual data,
so that each network specializes in a given modality. The visual subnetwork is
a pre-trained VGG16 model followed by a bidiretional recurrent layer, while the
residual audio subnetwork is based on stacked squeeze-excitation convolutional
blocks trained from scratch. After training each subnetwork, the fusion of
information from the audio and visual streams is performed at two different
stages. The early fusion stage combines features resulting from the last
convolutional block of the respective subnetworks at different time steps to
feed a bidirectional recurrent structure. The late fusion stage combines the
output of the early fusion stage with the independent predictions provided by
the two subnetworks, resulting in the final prediction. We evaluate the method
using the recently published TAU Audio-Visual Urban Scenes 2021, which contains
synchronized audio and video recordings from 12 European cities in 10 different
scene classes. The proposed model has been shown to provide an excellent
trade-off between prediction performance (86.5%) and system complexity (15M
parameters) in the evaluation results of the DCASE 2021 Challenge.
| [
{
"created": "Wed, 28 Jul 2021 06:10:10 GMT",
"version": "v1"
}
] | 2021-07-29 | [
[
"Naranjo-Alcazar",
"Javier",
""
],
[
"Perez-Castanos",
"Sergi",
""
],
[
"Lopez-Garcia",
"Aaron",
""
],
[
"Zuccarello",
"Pedro",
""
],
[
"Cobos",
"Maximo",
""
],
[
"Ferri",
"Francesc J.",
""
]
] | The use of multiple and semantically correlated sources can provide complementary information to each other that may not be evident when working with individual modalities on their own. In this context, multi-modal models can help producing more accurate and robust predictions in machine learning tasks where audio-visual data is available. This paper presents a multi-modal model for automatic scene classification that exploits simultaneously auditory and visual information. The proposed approach makes use of two separate networks which are respectively trained in isolation on audio and visual data, so that each network specializes in a given modality. The visual subnetwork is a pre-trained VGG16 model followed by a bidiretional recurrent layer, while the residual audio subnetwork is based on stacked squeeze-excitation convolutional blocks trained from scratch. After training each subnetwork, the fusion of information from the audio and visual streams is performed at two different stages. The early fusion stage combines features resulting from the last convolutional block of the respective subnetworks at different time steps to feed a bidirectional recurrent structure. The late fusion stage combines the output of the early fusion stage with the independent predictions provided by the two subnetworks, resulting in the final prediction. We evaluate the method using the recently published TAU Audio-Visual Urban Scenes 2021, which contains synchronized audio and video recordings from 12 European cities in 10 different scene classes. The proposed model has been shown to provide an excellent trade-off between prediction performance (86.5%) and system complexity (15M parameters) in the evaluation results of the DCASE 2021 Challenge. |
2009.04416 | Karl Cobbe | Karl Cobbe, Jacob Hilton, Oleg Klimov, John Schulman | Phasic Policy Gradient | null | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce Phasic Policy Gradient (PPG), a reinforcement learning framework
which modifies traditional on-policy actor-critic methods by separating policy
and value function training into distinct phases. In prior methods, one must
choose between using a shared network or separate networks to represent the
policy and value function. Using separate networks avoids interference between
objectives, while using a shared network allows useful features to be shared.
PPG is able to achieve the best of both worlds by splitting optimization into
two phases, one that advances training and one that distills features. PPG also
enables the value function to be more aggressively optimized with a higher
level of sample reuse. Compared to PPO, we find that PPG significantly improves
sample efficiency on the challenging Procgen Benchmark.
| [
{
"created": "Wed, 9 Sep 2020 16:52:53 GMT",
"version": "v1"
}
] | 2020-09-10 | [
[
"Cobbe",
"Karl",
""
],
[
"Hilton",
"Jacob",
""
],
[
"Klimov",
"Oleg",
""
],
[
"Schulman",
"John",
""
]
] | We introduce Phasic Policy Gradient (PPG), a reinforcement learning framework which modifies traditional on-policy actor-critic methods by separating policy and value function training into distinct phases. In prior methods, one must choose between using a shared network or separate networks to represent the policy and value function. Using separate networks avoids interference between objectives, while using a shared network allows useful features to be shared. PPG is able to achieve the best of both worlds by splitting optimization into two phases, one that advances training and one that distills features. PPG also enables the value function to be more aggressively optimized with a higher level of sample reuse. Compared to PPO, we find that PPG significantly improves sample efficiency on the challenging Procgen Benchmark. |
1303.5855 | Zhong-Yuan Zhang | Zhong-Yuan Zhang and Yong Wang and Yong-Yeol Ahn | Overlapping Community Detection in Complex Networks using Symmetric
Binary Matrix Factorization | null | null | 10.1103/PhysRevE.87.062803 | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Discovering overlapping community structures is a crucial step to
understanding the structure and dynamics of many networks. In this paper we
develop a symmetric binary matrix factorization model (SBMF) to identify
overlapping communities. Our model allows us not only to assign community
memberships explicitly to nodes, but also to distinguish outliers from
overlapping nodes. In addition, we propose a modified partition density to
evaluate the quality of community structures. We use this to determine the most
appropriate number of communities. We evaluate our methods using both synthetic
benchmarks and real world networks, demonstrating the effectiveness of our
approach.
| [
{
"created": "Sat, 23 Mar 2013 15:16:44 GMT",
"version": "v1"
}
] | 2015-06-15 | [
[
"Zhang",
"Zhong-Yuan",
""
],
[
"Wang",
"Yong",
""
],
[
"Ahn",
"Yong-Yeol",
""
]
] | Discovering overlapping community structures is a crucial step to understanding the structure and dynamics of many networks. In this paper we develop a symmetric binary matrix factorization model (SBMF) to identify overlapping communities. Our model allows us not only to assign community memberships explicitly to nodes, but also to distinguish outliers from overlapping nodes. In addition, we propose a modified partition density to evaluate the quality of community structures. We use this to determine the most appropriate number of communities. We evaluate our methods using both synthetic benchmarks and real world networks, demonstrating the effectiveness of our approach. |
0909.4767 | Christine Bachoc | Christine Bachoc (IMB) | Semidefinite programming, harmonic analysis and coding theory | null | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | These lecture notes where presented as a course of the CIMPA summer school in
Manila, July 20-30, 2009, Semidefinite programming in algebraic combinatorics.
This version is an update June 2010.
| [
{
"created": "Fri, 25 Sep 2009 19:04:18 GMT",
"version": "v1"
},
{
"created": "Wed, 8 Sep 2010 11:37:01 GMT",
"version": "v2"
}
] | 2010-09-09 | [
[
"Bachoc",
"Christine",
"",
"IMB"
]
] | These lecture notes where presented as a course of the CIMPA summer school in Manila, July 20-30, 2009, Semidefinite programming in algebraic combinatorics. This version is an update June 2010. |
1401.3075 | Qifu Sun | Qifu (Tyler) Sun, Xunrui Yin, Zongpeng Li, and Keping Long | Multicast Network Coding and Field Sizes | null | null | 10.1109/TIT.2015.2473863 | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In an acyclic multicast network, it is well known that a linear network
coding solution over GF($q$) exists when $q$ is sufficiently large. In
particular, for each prime power $q$ no smaller than the number of receivers, a
linear solution over GF($q$) can be efficiently constructed. In this work, we
reveal that a linear solution over a given finite field does \emph{not}
necessarily imply the existence of a linear solution over all larger finite
fields. Specifically, we prove by construction that: (i) For every source
dimension no smaller than 3, there is a multicast network linearly solvable
over GF(7) but not over GF(8), and another multicast network linearly solvable
over GF(16) but not over GF(17); (ii) There is a multicast network linearly
solvable over GF(5) but not over such GF($q$) that $q > 5$ is a Mersenne prime
plus 1, which can be extremely large; (iii) A multicast network linearly
solvable over GF($q^{m_1}$) and over GF($q^{m_2}$) is \emph{not} necessarily
linearly solvable over GF($q^{m_1+m_2}$); (iv) There exists a class of
multicast networks with a set $T$ of receivers such that the minimum field size
$q_{min}$ for a linear solution over GF($q_{min}$) is lower bounded by
$\Theta(\sqrt{|T|})$, but not every larger field than GF($q_{min}$) suffices to
yield a linear solution. The insight brought from this work is that not only
the field size, but also the order of subgroups in the multiplicative group of
a finite field affects the linear solvability of a multicast network.
| [
{
"created": "Tue, 14 Jan 2014 06:03:54 GMT",
"version": "v1"
},
{
"created": "Fri, 13 Feb 2015 06:34:47 GMT",
"version": "v2"
}
] | 2016-09-26 | [
[
"Qifu",
"",
"",
"Tyler"
],
[
"Sun",
"",
""
],
[
"Yin",
"Xunrui",
""
],
[
"Li",
"Zongpeng",
""
],
[
"Long",
"Keping",
""
]
] | In an acyclic multicast network, it is well known that a linear network coding solution over GF($q$) exists when $q$ is sufficiently large. In particular, for each prime power $q$ no smaller than the number of receivers, a linear solution over GF($q$) can be efficiently constructed. In this work, we reveal that a linear solution over a given finite field does \emph{not} necessarily imply the existence of a linear solution over all larger finite fields. Specifically, we prove by construction that: (i) For every source dimension no smaller than 3, there is a multicast network linearly solvable over GF(7) but not over GF(8), and another multicast network linearly solvable over GF(16) but not over GF(17); (ii) There is a multicast network linearly solvable over GF(5) but not over such GF($q$) that $q > 5$ is a Mersenne prime plus 1, which can be extremely large; (iii) A multicast network linearly solvable over GF($q^{m_1}$) and over GF($q^{m_2}$) is \emph{not} necessarily linearly solvable over GF($q^{m_1+m_2}$); (iv) There exists a class of multicast networks with a set $T$ of receivers such that the minimum field size $q_{min}$ for a linear solution over GF($q_{min}$) is lower bounded by $\Theta(\sqrt{|T|})$, but not every larger field than GF($q_{min}$) suffices to yield a linear solution. The insight brought from this work is that not only the field size, but also the order of subgroups in the multiplicative group of a finite field affects the linear solvability of a multicast network. |
2112.09631 | Archan Ray | Archan Ray, Nicholas Monath, Andrew McCallum, Cameron Musco | Sublinear Time Approximation of Text Similarity Matrices | 25 pages, 10 figures | null | null | null | cs.LG cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We study algorithms for approximating pairwise similarity matrices that arise
in natural language processing. Generally, computing a similarity matrix for
$n$ data points requires $\Omega(n^2)$ similarity computations. This quadratic
scaling is a significant bottleneck, especially when similarities are computed
via expensive functions, e.g., via transformer models. Approximation methods
reduce this quadratic complexity, often by using a small subset of exactly
computed similarities to approximate the remainder of the complete pairwise
similarity matrix.
Significant work focuses on the efficient approximation of positive
semidefinite (PSD) similarity matrices, which arise e.g., in kernel methods.
However, much less is understood about indefinite (non-PSD) similarity
matrices, which often arise in NLP. Motivated by the observation that many of
these matrices are still somewhat close to PSD, we introduce a generalization
of the popular Nystr\"{o}m method to the indefinite setting. Our algorithm can
be applied to any similarity matrix and runs in sublinear time in the size of
the matrix, producing a rank-$s$ approximation with just $O(ns)$ similarity
computations.
We show that our method, along with a simple variant of CUR decomposition,
performs very well in approximating a variety of similarity matrices arising in
NLP tasks. We demonstrate high accuracy of the approximated similarity matrices
in the downstream tasks of document classification, sentence similarity, and
cross-document coreference.
| [
{
"created": "Fri, 17 Dec 2021 17:04:34 GMT",
"version": "v1"
},
{
"created": "Wed, 16 Feb 2022 19:18:56 GMT",
"version": "v2"
},
{
"created": "Wed, 27 Apr 2022 13:56:51 GMT",
"version": "v3"
}
] | 2022-04-28 | [
[
"Ray",
"Archan",
""
],
[
"Monath",
"Nicholas",
""
],
[
"McCallum",
"Andrew",
""
],
[
"Musco",
"Cameron",
""
]
] | We study algorithms for approximating pairwise similarity matrices that arise in natural language processing. Generally, computing a similarity matrix for $n$ data points requires $\Omega(n^2)$ similarity computations. This quadratic scaling is a significant bottleneck, especially when similarities are computed via expensive functions, e.g., via transformer models. Approximation methods reduce this quadratic complexity, often by using a small subset of exactly computed similarities to approximate the remainder of the complete pairwise similarity matrix. Significant work focuses on the efficient approximation of positive semidefinite (PSD) similarity matrices, which arise e.g., in kernel methods. However, much less is understood about indefinite (non-PSD) similarity matrices, which often arise in NLP. Motivated by the observation that many of these matrices are still somewhat close to PSD, we introduce a generalization of the popular Nystr\"{o}m method to the indefinite setting. Our algorithm can be applied to any similarity matrix and runs in sublinear time in the size of the matrix, producing a rank-$s$ approximation with just $O(ns)$ similarity computations. We show that our method, along with a simple variant of CUR decomposition, performs very well in approximating a variety of similarity matrices arising in NLP tasks. We demonstrate high accuracy of the approximated similarity matrices in the downstream tasks of document classification, sentence similarity, and cross-document coreference. |
2406.17826 | Krzysztof Kotowski PhD | Krzysztof Kotowski, Christoph Haskamp, Jacek Andrzejewski, Bogdan
Ruszczak, Jakub Nalepa, Daniel Lakey, Peter Collins, Aybike Kolmas, Mauro
Bartesaghi, Jose Martinez-Heras, Gabriele De Canio | European Space Agency Benchmark for Anomaly Detection in Satellite
Telemetry | 87 pages, 24 figures, 19 tables | null | null | null | cs.LG cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Machine learning has vast potential to improve anomaly detection in satellite
telemetry which is a crucial task for spacecraft operations. This potential is
currently hampered by a lack of comprehensible benchmarks for multivariate time
series anomaly detection, especially for the challenging case of satellite
telemetry. The European Space Agency Benchmark for Anomaly Detection in
Satellite Telemetry (ESA-ADB) aims to address this challenge and establish a
new standard in the domain. It is a result of close cooperation between
spacecraft operations engineers from the European Space Agency (ESA) and
machine learning experts. The newly introduced ESA Anomalies Dataset contains
annotated real-life telemetry from three different ESA missions, out of which
two are included in ESA-ADB. Results of typical anomaly detection algorithms
assessed in our novel hierarchical evaluation pipeline show that new approaches
are necessary to address operators' needs. All elements of ESA-ADB are publicly
available to ensure its full reproducibility.
| [
{
"created": "Tue, 25 Jun 2024 13:23:37 GMT",
"version": "v1"
}
] | 2024-06-27 | [
[
"Kotowski",
"Krzysztof",
""
],
[
"Haskamp",
"Christoph",
""
],
[
"Andrzejewski",
"Jacek",
""
],
[
"Ruszczak",
"Bogdan",
""
],
[
"Nalepa",
"Jakub",
""
],
[
"Lakey",
"Daniel",
""
],
[
"Collins",
"Peter",
""
],
[
"Kolmas",
"Aybike",
""
],
[
"Bartesaghi",
"Mauro",
""
],
[
"Martinez-Heras",
"Jose",
""
],
[
"De Canio",
"Gabriele",
""
]
] | Machine learning has vast potential to improve anomaly detection in satellite telemetry which is a crucial task for spacecraft operations. This potential is currently hampered by a lack of comprehensible benchmarks for multivariate time series anomaly detection, especially for the challenging case of satellite telemetry. The European Space Agency Benchmark for Anomaly Detection in Satellite Telemetry (ESA-ADB) aims to address this challenge and establish a new standard in the domain. It is a result of close cooperation between spacecraft operations engineers from the European Space Agency (ESA) and machine learning experts. The newly introduced ESA Anomalies Dataset contains annotated real-life telemetry from three different ESA missions, out of which two are included in ESA-ADB. Results of typical anomaly detection algorithms assessed in our novel hierarchical evaluation pipeline show that new approaches are necessary to address operators' needs. All elements of ESA-ADB are publicly available to ensure its full reproducibility. |
2207.01234 | Vishnu Raj | Vishnu Raj, Tianyu Cui, Markus Heinonen and Pekka Marttinen | Incorporating functional summary information in Bayesian neural networks
using a Dirichlet process likelihood approach | Accepted in AISTATS 2023 | null | null | null | cs.LG cs.AI stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Bayesian neural networks (BNNs) can account for both aleatoric and epistemic
uncertainty. However, in BNNs the priors are often specified over the weights
which rarely reflects true prior knowledge in large and complex neural network
architectures. We present a simple approach to incorporate prior knowledge in
BNNs based on external summary information about the predicted classification
probabilities for a given dataset. The available summary information is
incorporated as augmented data and modeled with a Dirichlet process, and we
derive the corresponding \emph{Summary Evidence Lower BOund}. The approach is
founded on Bayesian principles, and all hyperparameters have a proper
probabilistic interpretation. We show how the method can inform the model about
task difficulty and class imbalance. Extensive experiments show that, with
negligible computational overhead, our method parallels and in many cases
outperforms popular alternatives in accuracy, uncertainty calibration, and
robustness against corruptions with both balanced and imbalanced data.
| [
{
"created": "Mon, 4 Jul 2022 07:06:45 GMT",
"version": "v1"
},
{
"created": "Tue, 24 Jan 2023 08:08:11 GMT",
"version": "v2"
}
] | 2023-01-25 | [
[
"Raj",
"Vishnu",
""
],
[
"Cui",
"Tianyu",
""
],
[
"Heinonen",
"Markus",
""
],
[
"Marttinen",
"Pekka",
""
]
] | Bayesian neural networks (BNNs) can account for both aleatoric and epistemic uncertainty. However, in BNNs the priors are often specified over the weights which rarely reflects true prior knowledge in large and complex neural network architectures. We present a simple approach to incorporate prior knowledge in BNNs based on external summary information about the predicted classification probabilities for a given dataset. The available summary information is incorporated as augmented data and modeled with a Dirichlet process, and we derive the corresponding \emph{Summary Evidence Lower BOund}. The approach is founded on Bayesian principles, and all hyperparameters have a proper probabilistic interpretation. We show how the method can inform the model about task difficulty and class imbalance. Extensive experiments show that, with negligible computational overhead, our method parallels and in many cases outperforms popular alternatives in accuracy, uncertainty calibration, and robustness against corruptions with both balanced and imbalanced data. |
2210.07564 | Xin Tian | Xin Tian, Yingzhan Lin, Mengfei Song, Siqi Bao, Fan Wang, Huang He,
Shuqi Sun, Hua Wu | Q-TOD: A Query-driven Task-oriented Dialogue System | Accepted to EMNLP 2022 | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Existing pipelined task-oriented dialogue systems usually have difficulties
adapting to unseen domains, whereas end-to-end systems are plagued by
large-scale knowledge bases in practice. In this paper, we introduce a novel
query-driven task-oriented dialogue system, namely Q-TOD. The essential
information from the dialogue context is extracted into a query, which is
further employed to retrieve relevant knowledge records for response
generation. Firstly, as the query is in the form of natural language and not
confined to the schema of the knowledge base, the issue of domain adaption is
alleviated remarkably in Q-TOD. Secondly, as the query enables the decoupling
of knowledge retrieval from the generation, Q-TOD gets rid of the issue of
knowledge base scalability. To evaluate the effectiveness of the proposed
Q-TOD, we collect query annotations for three publicly available task-oriented
dialogue datasets. Comprehensive experiments verify that Q-TOD outperforms
strong baselines and establishes a new state-of-the-art performance on these
datasets.
| [
{
"created": "Fri, 14 Oct 2022 06:38:19 GMT",
"version": "v1"
}
] | 2022-10-17 | [
[
"Tian",
"Xin",
""
],
[
"Lin",
"Yingzhan",
""
],
[
"Song",
"Mengfei",
""
],
[
"Bao",
"Siqi",
""
],
[
"Wang",
"Fan",
""
],
[
"He",
"Huang",
""
],
[
"Sun",
"Shuqi",
""
],
[
"Wu",
"Hua",
""
]
] | Existing pipelined task-oriented dialogue systems usually have difficulties adapting to unseen domains, whereas end-to-end systems are plagued by large-scale knowledge bases in practice. In this paper, we introduce a novel query-driven task-oriented dialogue system, namely Q-TOD. The essential information from the dialogue context is extracted into a query, which is further employed to retrieve relevant knowledge records for response generation. Firstly, as the query is in the form of natural language and not confined to the schema of the knowledge base, the issue of domain adaption is alleviated remarkably in Q-TOD. Secondly, as the query enables the decoupling of knowledge retrieval from the generation, Q-TOD gets rid of the issue of knowledge base scalability. To evaluate the effectiveness of the proposed Q-TOD, we collect query annotations for three publicly available task-oriented dialogue datasets. Comprehensive experiments verify that Q-TOD outperforms strong baselines and establishes a new state-of-the-art performance on these datasets. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.