aid stringlengths 9 15 | mid stringlengths 7 10 | abstract stringlengths 78 2.56k | related_work stringlengths 92 1.77k | ref_abstract dict |
|---|---|---|---|---|
1901.06268 | 2908914784 | Biological data are extremely diverse, complex but also quite sparse. The recent developments in deep learning methods are offering new possibilities for the analysis of complex data. However, it is easy to be get a deep learning model that seems to have good results but is in fact either overfitting the training data ... | in @cite_9 are using stack auto-encoders to extract features from protein sequences. The classification predicting protein-protein interaction is then done by directly linking the output of the last auto-encoder to a softmax classifier. For their inputs, there are converting the sequences into fixed-size Boolean vector... | {
"cite_N": [
"@cite_9"
],
"mid": [
"2618265628"
],
"abstract": [
"Abstract Background Protein-protein interactions (PPIs) are critical for many biological processes. It is therefore important to develop accurate high-throughput methods for identifying PPI to better understand protein function, di... |
1901.06268 | 2908914784 | Biological data are extremely diverse, complex but also quite sparse. The recent developments in deep learning methods are offering new possibilities for the analysis of complex data. However, it is easy to be get a deep learning model that seems to have good results but is in fact either overfitting the training data ... | proposed in @cite_0 a plain fully connected neural network, similar to our first model in this paper but significantly bigger, with layers containing 512, 256 and 128 units for what should be feature extraction, and 128 units for the head of their network. However, they are not giving as input to the network protein se... | {
"cite_N": [
"@cite_0"
],
"mid": [
"2616246685"
],
"abstract": [
"The complex language of eukaryotic gene expression remains incompletely understood. Despite the importance suggested by many proteins variants statistically associated with human disease, nearly all such variants have unknown mecha... |
1901.06268 | 2908914784 | Biological data are extremely diverse, complex but also quite sparse. The recent developments in deep learning methods are offering new possibilities for the analysis of complex data. However, it is easy to be get a deep learning model that seems to have good results but is in fact either overfitting the training data ... | use in @cite_17 a Deep Polynomial Network on features extracted by hand, like amino acid mutation rates or hydrophobic properties of proteins, to make their classification. Thus, they do not use the chain of amino acid residues as an input. They based the learning process on a 5-fold cross validation without test sets. | {
"cite_N": [
"@cite_17"
],
"mid": [
"2808071711"
],
"abstract": [
"Predicting the protein–protein interactions (PPIs) has played an important role in many applications. Hence, a novel computational method for PPIs prediction is highly desirable. PPIs endow with protein amino acid mutation rate an... |
1901.06268 | 2908914784 | Biological data are extremely diverse, complex but also quite sparse. The recent developments in deep learning methods are offering new possibilities for the analysis of complex data. However, it is easy to be get a deep learning model that seems to have good results but is in fact either overfitting the training data ... | @cite_24 , present a model composed of an embedding layer, three convolutions and a LSTM layer for feature extractions of protein sequences, before concatenating LSTM output of both proteins and performing classification with a fully connected layer linked to a sigmoid classifier. The architecture of our recurrent mode... | {
"cite_N": [
"@cite_24"
],
"mid": [
"2885583144"
],
"abstract": [
"Machine learning based predictions of protein–protein interactions (PPIs) could provide valuable insights into protein functions, disease occurrence, and therapy design on a large scale. The intensive feature engineering in most o... |
1901.06268 | 2908914784 | Biological data are extremely diverse, complex but also quite sparse. The recent developments in deep learning methods are offering new possibilities for the analysis of complex data. However, it is easy to be get a deep learning model that seems to have good results but is in fact either overfitting the training data ... | Finally, @cite_10 present a fully connected model regulated by dropouts. Like @cite_0 , they use composition-transition-distribution descriptors as features. They apply a 5-fold cross validation and have no separated test sets. | {
"cite_N": [
"@cite_0",
"@cite_10"
],
"mid": [
"2616246685",
"2804331675"
],
"abstract": [
"The complex language of eukaryotic gene expression remains incompletely understood. Despite the importance suggested by many proteins variants statistically associated with human disease, nearly al... |
1901.06268 | 2908914784 | Biological data are extremely diverse, complex but also quite sparse. The recent developments in deep learning methods are offering new possibilities for the analysis of complex data. However, it is easy to be get a deep learning model that seems to have good results but is in fact either overfitting the training data ... | We can also mentioned the work of @cite_8 , proposing a multi-layered LSTM model to predict interface residue pair interactions, thus at a finer level level than prediction interaction between two proteins. This is a direction towards which we would like to extend our results. | {
"cite_N": [
"@cite_8"
],
"mid": [
"2614995786"
],
"abstract": [
"Motivation: Proteins usually fulfill their biological functions by interacting with other proteins. Although some methods have been developed to predict the binding sites of a monomer protein, these are not sufficient for predictio... |
1901.06263 | 2911010497 | Considering the advances in building monitoring and control through networks of interconnected devices, effective handling of the associated rich data streams is becoming an important challenge. In many situations the application of conventional system identification or approximate grey-box models, partly theoretic and... | A paper focused on energy-efficiency improvements leveraging available building-level data for data mining is @cite_7 . The authors list the main predictive tasks in which data mining of large quantities of measurements and contextual information is relevant. These cover: building energy demand prediction, building occ... | {
"cite_N": [
"@cite_5",
"@cite_7",
"@cite_17"
],
"mid": [
"2801774872",
"2204655733",
"2754029504"
],
"abstract": [
"Abstract Due to its significant contribution to global energy usage and the associated greenhouse gas emissions, existing building stock's energy efficiency must im... |
1901.06263 | 2911010497 | Considering the advances in building monitoring and control through networks of interconnected devices, effective handling of the associated rich data streams is becoming an important challenge. In many situations the application of conventional system identification or approximate grey-box models, partly theoretic and... | Deployment of distributed sensor networks for finer grained spatio-temporal monitoring of indoor conditions is performed by @cite_13 . The authors argue that the statistical modelling of the indoor environment as non-parametric Gaussian processes can lead to reliable information that is fed back to the building managem... | {
"cite_N": [
"@cite_24",
"@cite_18",
"@cite_13",
"@cite_22"
],
"mid": [
"2804138105",
"2594724848",
"2743217645",
"2755807005"
],
"abstract": [
"This paper presents an Internet of Things (IoT) platform for a smart building which provides human care services for occupants. ... |
1901.06263 | 2911010497 | Considering the advances in building monitoring and control through networks of interconnected devices, effective handling of the associated rich data streams is becoming an important challenge. In many situations the application of conventional system identification or approximate grey-box models, partly theoretic and... | As compared to traditional model-based control (MBC), data-driven control (DDC) represents an emerging field of study which accounts for the need to manage the data deluge produced by dense temporal and spatial monitoring of various systems. A broad survey on the specific nature of DDC and comparison to MBC in various ... | {
"cite_N": [
"@cite_0",
"@cite_16"
],
"mid": [
"2040871222",
"2790404719"
],
"abstract": [
"This paper is a brief survey on the existing problems and challenges inherent in model-based control (MBC) theory, and some important issues in the analysis and design of data-driven control (DDC) ... |
1901.06263 | 2911010497 | Considering the advances in building monitoring and control through networks of interconnected devices, effective handling of the associated rich data streams is becoming an important challenge. In many situations the application of conventional system identification or approximate grey-box models, partly theoretic and... | Big data analytics for smart city electricity consumption in presented in @cite_28 . The authors use computational intelligence algorithms to model the consumption of eight university buildings. The outcome consists of offline policies to optimise energy usage across the campus. In @cite_11 a different application is d... | {
"cite_N": [
"@cite_28",
"@cite_23",
"@cite_25",
"@cite_20",
"@cite_11"
],
"mid": [
"2790486829",
"",
"2294864664",
"2788535427",
"2475772748"
],
"abstract": [
"New technologies such as sensor networks have been incorporated into the management of buildings for org... |
1901.06263 | 2911010497 | Considering the advances in building monitoring and control through networks of interconnected devices, effective handling of the associated rich data streams is becoming an important challenge. In many situations the application of conventional system identification or approximate grey-box models, partly theoretic and... | @cite_2 describe in detail the explicit data modelling process for smart building evaluation. A case study is carried out for energy forecasting of a target building using techniques such a Bayesian Regularized Neural Networks and Random Forests. SVM are also considered but provide weaker results in this specific scena... | {
"cite_N": [
"@cite_14",
"@cite_2"
],
"mid": [
"1558927261",
"2364005411"
],
"abstract": [
"As our society gains a better understanding of how humans have negatively impacted the environment, research related to reducing carbon emissions and overall energy consumption has become increasin... |
1901.06263 | 2911010497 | Considering the advances in building monitoring and control through networks of interconnected devices, effective handling of the associated rich data streams is becoming an important challenge. In many situations the application of conventional system identification or approximate grey-box models, partly theoretic and... | The current paper also builds upon own previous work dedicated to decision support systems for renewable energy campus microgrids @cite_10 and carrying out Model Predictive Control (MPC) for building simulations @cite_3 . Earlier work has also included exploratory data analysis from a single building AHU without furthe... | {
"cite_N": [
"@cite_9",
"@cite_10",
"@cite_3"
],
"mid": [
"2770442362",
"2574708214",
"2512691281"
],
"abstract": [
"Modern, densely instrumented, smart buildings generate large amounts of raw data. This poses significant challenges from both the data management perspective as wel... |
1901.06261 | 2910933843 | Application of neural networks to a vast variety of practical applications is transforming the way AI is applied in practice. Pre-trained neural network models available through APIs or capability to custom train pre-built neural network architectures with customer data has made the consumption of AI by developers much... | Evolutionary algorithms and reinforcement learning are currently the two state-of-the-art techniques used by neural network architectures search algorithms. With Neural Architecture Search @cite_38 , demonstrated in an experiment over 28 days and with 800 GPUs that neural network architectures with performances close t... | {
"cite_N": [
"@cite_38",
"@cite_0",
"@cite_71",
"@cite_13",
"@cite_25"
],
"mid": [
"2963374479",
"2747469359",
"2964081807",
"2556833785",
"2773706593"
],
"abstract": [
"Neural networks are powerful and flexible models that work well for many difficult learning tas... |
1901.06261 | 2910933843 | Application of neural networks to a vast variety of practical applications is transforming the way AI is applied in practice. Pre-trained neural network models available through APIs or capability to custom train pre-built neural network architectures with customer data has made the consumption of AI by developers much... | Various techniques exist which try to shorten the training time. One idea is based on the idea of terminating unpromising training runs early. The partially observed learning curve is used directly to decide to terminate a run early @cite_73 or first extrapolated and then used @cite_82 @cite_61 @cite_21 . Other methods... | {
"cite_N": [
"@cite_61",
"@cite_28",
"@cite_21",
"@cite_65",
"@cite_52",
"@cite_19",
"@cite_49",
"@cite_73",
"@cite_25",
"@cite_82"
],
"mid": [
"2751836095",
"2771751675",
"",
"2785366763",
"2914037680",
"2963137684",
"2951104886",
"",
"... |
1901.06257 | 2910992039 | Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (... | Many studies on GPS trajectory mining exist, such as user activity estimation @cite_3 @cite_2 @cite_17 @cite_16 , transportation mode detection @cite_4 @cite_43 , and region analysis @cite_37 @cite_0 . A typical approach to tackle these tasks first extracts stay-points as a clue for solving them. Therefore, we believe ... | {
"cite_N": [
"@cite_37",
"@cite_4",
"@cite_3",
"@cite_0",
"@cite_43",
"@cite_2",
"@cite_16",
"@cite_17"
],
"mid": [
"2075190119",
"2143394441",
"2110953678",
"2153207204",
"578585133",
"1988545169",
"2171805028",
"2138460198"
],
"abstract": [
... |
1901.06257 | 2910992039 | Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (... | Various stay-point extraction methods have already been proposed. For example, Ashbrook and Starner @cite_13 @cite_19 use a modified @math -means method, @cite_20 use DBSCAN @cite_9 , and @cite_39 employ Mean-Shift @cite_24 , all of which are based on clustering. @cite_25 and @cite_26 assume that stay-points are positi... | {
"cite_N": [
"@cite_26",
"@cite_9",
"@cite_39",
"@cite_24",
"@cite_19",
"@cite_23",
"@cite_13",
"@cite_25",
"@cite_20"
],
"mid": [
"2140251882",
"1673310716",
"2072787293",
"2022686119",
"2009155608",
"2779280457",
"2168685189",
"2071049788",
... |
1901.06257 | 2910992039 | Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (... | The challenge that most resembles our personalized visited-POI assignment task is detecting semantic locations from GPS trajectory data @cite_22 @cite_28 @cite_26 . @cite_22 extracted stay-points from trajectories and combined them with street addresses obtained by a reverse geocoder. Their method assigns a semantic la... | {
"cite_N": [
"@cite_28",
"@cite_26",
"@cite_22"
],
"mid": [
"2122338787",
"2140251882",
"2067193733"
],
"abstract": [
"With help of context, computer systems and applications could be more user-friendly, flexible and adaptable. With semantic locations, applications can understand ... |
1901.06257 | 2910992039 | Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (... | POI recommendation tasks are closely related to our target task. Many previous studies have addressed POI recommendations @cite_29 @cite_31 @cite_14 @cite_27 @cite_33 . Most used the collaborative filtering (CF) approach, which requires inter-user information, to achieve recommendations. @cite_31 performed co-clusterin... | {
"cite_N": [
"@cite_14",
"@cite_33",
"@cite_29",
"@cite_27",
"@cite_31"
],
"mid": [
"2112631146",
"2073013176",
"2139809240",
"2087692915",
"2009799282"
],
"abstract": [
"This paper studies the problem of recommending new venues to users who participate in location... |
1901.06257 | 2910992039 | Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (... | Other studies on a location naming task @cite_18 @cite_36 and a POI recommendation task @cite_15 use a supervised learning algorithm @cite_10 @cite_44 to build POI ranking models. To formalize their problems as a ranking challenge, they look at a location (i.e., longitude and latitude) as a query and a user's check-in ... | {
"cite_N": [
"@cite_18",
"@cite_22",
"@cite_36",
"@cite_44",
"@cite_15",
"@cite_10"
],
"mid": [
"2082576493",
"2067193733",
"2029550988",
"2149427297",
"2026532078",
"2108862644"
],
"abstract": [
"Many innovative location-based services have been establishe... |
1901.06257 | 2910992039 | Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (... | More recently, several novel tasks have been proposed that are related to visited-POI assignments. For example, @cite_34 characterized the life cycle of POIs and investigated the POI evolution process over time. Espin- @cite_6 tackled a task that clusters users based on spatio-temporal dimensions with a non-negative te... | {
"cite_N": [
"@cite_34",
"@cite_6"
],
"mid": [
"2507653743",
"2338768639"
],
"abstract": [
"A Point of Interest (POI) refers to a specific location that people may find useful or interesting. While a large body of research has been focused on identifying and recommending POIs, there are f... |
1901.06257 | 2910992039 | Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (... | @cite_41 focused on the periodic behaviors of users and formalized POI check-in patterns as a stochastic point process. An interesting aspect of their method is that they take into account a factor of the influence of the close friends of users. In contrast, our task detects actual visited-POIs from obtained raw GPS tr... | {
"cite_N": [
"@cite_41"
],
"mid": [
"2549560474"
],
"abstract": [
"Social networks are getting closer to our real physical world. People share the exact location and time of their check-ins and are influenced by their friends. Modeling the spatio-temporal behavior of users in social networks is o... |
1901.06257 | 2910992039 | Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (... | @cite_30 proposed a sequential personalized spatial item recommendation framework (SPORE), which recommends a sequence of POIs based on individual POI-visit histories. Their target closely resembles ours. However, the essential difference is that their task assumes a sequence of check-in records as input, unlike raw GP... | {
"cite_N": [
"@cite_30"
],
"mid": [
"2434565296"
],
"abstract": [
"With the rapid development of location-based social networks (LBSNs), spatial item recommendation has become an important way of helping users discover interesting locations to increase their engagement with location-based service... |
1901.06257 | 2910992039 | Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (... | @cite_32 proposed a task that detects personally semantic places from GPS trajectories. Their proposed task also appears to closely resemble ours. However, their target is to detect places ( frequently visited by an individual user) that might have such important semantic meanings as home or office. In this perspective... | {
"cite_N": [
"@cite_41",
"@cite_32"
],
"mid": [
"2549560474",
"2039001267"
],
"abstract": [
"Social networks are getting closer to our real physical world. People share the exact location and time of their check-ins and are influenced by their friends. Modeling the spatio-temporal behavio... |
1901.06257 | 2910992039 | Knowledge discovery from GPS trajectory data is an important topic in several scientific areas, including data mining, human behavior analysis, and user modeling. This paper proposes a task that assigns personalized visited-POIs. Its goal is to estimate fine-grained and pre-defined locations (i.e., points of interest (... | @cite_40 employed a Bayesian network to detect the categories of visited-POIs, such as hospitals and universities, from the GPS trajectories of vehicles. Their motivation is closely related to ours. The essential difference is that they only detect the categories of visited-POIs; we detect the visited-POIs themselves. ... | {
"cite_N": [
"@cite_40"
],
"mid": [
"2614175038"
],
"abstract": [
"Identifying visited points of interest (PoIs) from vehicle trajectories remains an open problem that is difficult due to vehicles parking often at some distance from the visited PoI and due to some regions having a high PoI densit... |
1907.08015 | 2956604637 | The evolution and development of events have their own basic principles, which make events happen sequentially. Therefore, the discovery of such evolutionary patterns among events are of great value for event prediction, decision-making and scenario design of dialog systems. However, conventional knowledge graph mainly... | The most relevant research area with ELG is script learning. The use of scripts in AI dates back to the 1970s @cite_4 @cite_17 . In this study, are an influential early encoding of situation-specific world event. In recent years, a growing body of research has investigated statistical script learning. , proposed unsupe... | {
"cite_N": [
"@cite_14",
"@cite_4",
"@cite_17"
],
"mid": [
"2145374219",
"2000900121",
"2142269169"
],
"abstract": [
"Scripts represent knowledge of stereotypical event sequences that can aid text understanding. Initial statistical methods have been developed to learn probabilisti... |
1907.07826 | 2959681520 | Detecting emotions from text is an extension of simple sentiment polarity detection. Instead of considering only positive or negative sentiments, emotions are conveyed using more tangible manner; thus, they can be expressed as many shades of gray. This paper manifests the results of our experimentation for fine-grained... | In a different paper @cite_0 , the authors described the preparation of the Bengali WordNet Affect containing six types of emotion words. They employed an automatic method of sense disambiguation. The Bengali WordNet Affect could be useful for emotion-related language processing tasks in Bengali. | {
"cite_N": [
"@cite_0"
],
"mid": [
"2065586386"
],
"abstract": [
"The present discussion highlights the aspects of an ongoing doctoral thesis grounded on the analysis and tracking of emotions from English and Bengali texts. Development of lexical resources and corpora meets the preliminary urgenc... |
1907.07826 | 2959681520 | Detecting emotions from text is an extension of simple sentiment polarity detection. Instead of considering only positive or negative sentiments, emotions are conveyed using more tangible manner; thus, they can be expressed as many shades of gray. This paper manifests the results of our experimentation for fine-grained... | On a case study for Bengali @cite_4 , the authors considered 1,100 sentences on eight different topics. They prepared a knowledge base for emoticons and also employed a morphological analyzer to identify the lexical keywords from the Bengali WordNet Affect lists. They claimed an overall precision, recall and F1-Score (... | {
"cite_N": [
"@cite_4"
],
"mid": [
"193025648"
],
"abstract": [
"Rapid growth of blogs in the Web 2.0 and the handshaking between multilingual search and sentiment analysis motivate us to develop a blog based emotion analysis system for Bengali. The present paper describes the identification, vis... |
1907.07885 | 2960548256 | We introduce a formal framework for analyzing trades in financial markets. An exchange is where multiple buyers and sellers participate to trade. These days, all big exchanges use computer algorithms that implement double sided auctions to match buy and sell requests and these algorithms must abide by certain regulator... | There is no prior work known to us which formalizes financial algorithms used by the exchanges. Passmore and Ignatovich in @cite_17 highlight the significance, opportunities and challenges involved in formalizing financial markets. Their work describes in detail the whole spectrum of financial algorithms that need to b... | {
"cite_N": [
"@cite_17"
],
"mid": [
"2735447866"
],
"abstract": [
"Many deep issues plaguing today’s financial markets are symptoms of a fundamental problem: The complexity of algorithms underlying modern finance has significantly outpaced the power of traditional tools used to design and regulat... |
1907.07885 | 2960548256 | We introduce a formal framework for analyzing trades in financial markets. An exchange is where multiple buyers and sellers participate to trade. These days, all big exchanges use computer algorithms that implement double sided auctions to match buy and sell requests and these algorithms must abide by certain regulator... | On the other hand, there are quite a few works formalizing various concepts from auction theory @cite_15 @cite_6 @cite_7 . Most of these works focus on the Vickrey auction mechanism. In Vickrey auction, there is a single seller with different items and multiple buyers with valuations for every subsets of items. Each bu... | {
"cite_N": [
"@cite_15",
"@cite_7",
"@cite_6"
],
"mid": [
"2059840483",
"1796295358",
"1521892088"
],
"abstract": [
"We introduce formal methods' of mechanized reasoning from computer science to address two problems in auction design and practice: is a given auction design soundly... |
1907.07729 | 2961115932 | We consider the problem of rigid registration, where we wish to jointly register multiple point sets via rigid transforms. This arises in applications such as sensor network localization, multiview registration, and protein structure determination. The least-squares estimator for this problem can be reduced to a rank-c... | The rank-restricted subset @math of the PSD cone is nonconvex, which implies that standard convergence result for ADMM @cite_27 does not directly apply to . However, we do leverage the convergence of convex ADMM for analyzing the convergence of when the noise is low. A phase transition phenomena similar to the one cite... | {
"cite_N": [
"@cite_27",
"@cite_6"
],
"mid": [
"2164278908",
"2161367205"
],
"abstract": [
"Many problems of recent interest in statistics and machine learning can be posed in the framework of convex optimization. Due to the explosion in size and complexity of modern datasets, it is incre... |
1907.07729 | 2961115932 | We consider the problem of rigid registration, where we wish to jointly register multiple point sets via rigid transforms. This arises in applications such as sensor network localization, multiview registration, and protein structure determination. The least-squares estimator for this problem can be reduced to a rank-c... | The theoretical convergence of ADMM for nonconvex problems has been studied in @cite_17 @cite_0 @cite_7 . However, a crucial working assumption common to these results does not hold in our case. More precisely, observe that we can rewrite as where @math is the indicator function associated with a feasible set @math @ci... | {
"cite_N": [
"@cite_0",
"@cite_27",
"@cite_7",
"@cite_17"
],
"mid": [
"2155723734",
"2164278908",
"2962853966",
"2295652899"
],
"abstract": [
"Aiming at solving large-scale optimization problems, this paper studies distributed optimization methods based on the alternating ... |
1907.07729 | 2961115932 | We consider the problem of rigid registration, where we wish to jointly register multiple point sets via rigid transforms. This arises in applications such as sensor network localization, multiview registration, and protein structure determination. The least-squares estimator for this problem can be reduced to a rank-c... | We do not make such smoothness assumptions in our analysis. We can afford to do this since we are analyzing a special class of problems, as opposed to the more general setups in @cite_17 @cite_0 @cite_7 . Instead of showing a monotonic decrease in the augmented Lagrangian, our analysis relies on the phenomenon of tight... | {
"cite_N": [
"@cite_0",
"@cite_7",
"@cite_17"
],
"mid": [
"2155723734",
"2962853966",
"2295652899"
],
"abstract": [
"Aiming at solving large-scale optimization problems, this paper studies distributed optimization methods based on the alternating direction method of multipliers (A... |
1907.07803 | 2959891429 | One problem when studying how to find and fix syntax errors is how to get natural and representative examples of syntax errors. Most syntax error datasets are not free, open, and public, or they are extracted from novice programmers and do not represent syntax errors that the general population of developers would make... | Syntactically incorrect code is artificially derivable, as formal programming languages provide grammar rules which can be referred to for correctness. Random token level insertions, deletions, and replacements were performed to generate syntax errors from existing open source Java projects @cite_2 . 10.7287 peerj.prep... | {
"cite_N": [
"@cite_10",
"@cite_13",
"@cite_2"
],
"mid": [
"2135937266",
"2281672363",
"2060384944"
],
"abstract": [
"A good test suite is one that detects real faults. Because the set of faults in a program is usually unknowable, this definition is not useful to practitioners who... |
1907.07803 | 2959891429 | One problem when studying how to find and fix syntax errors is how to get natural and representative examples of syntax errors. Most syntax error datasets are not free, open, and public, or they are extracted from novice programmers and do not represent syntax errors that the general population of developers would make... | Automated source code repair, like identifying and refactoring improper method names, also required a labeled dataset of valid and invalid source code @cite_0 . Program repair is often viewed as different than syntax error correction because testing is performed which serves as a benchmark for repaired code, while synt... | {
"cite_N": [
"@cite_0"
],
"mid": [
"2943748428"
],
"abstract": [
"To ensure code readability and facilitate software maintenance, program methods must be named properly. In particular, method names must be consistent with the corresponding method implementations. Debugging method names remains an... |
1907.07803 | 2959891429 | One problem when studying how to find and fix syntax errors is how to get natural and representative examples of syntax errors. Most syntax error datasets are not free, open, and public, or they are extracted from novice programmers and do not represent syntax errors that the general population of developers would make... | Free and open datasets of naturally made errors and their fixes are more difficult to obtain. Blackbox, a data collection project within the BlueJ Java development environment, requires manual staff contact for access to data and forbids the release of the raw dataset @cite_11 . Pritchard:2015:FDE:2846680.2846681 analy... | {
"cite_N": [
"@cite_4",
"@cite_12",
"@cite_11"
],
"mid": [
"",
"2963074569",
"2057833160"
],
"abstract": [
"",
"Which programming error messages are the most common? We investigate this question, motivated by writing error explanations for novices. We consider large data sets ... |
1907.07769 | 2959758584 | We present a voice conversion solution using recurrent sequence to sequence modeling for DNNs. Our solution takes advantage of recent advances in attention based modeling in the fields of Neural Machine Translation (NMT), Text-to-Speech (TTS) and Automatic Speech Recognition (ASR). The problem consists of converting be... | Pertinent to our discussion are seq2seq modeling works @cite_42 @cite_16 . In these works, additional loss terms are introduced to encourage the model to learn alignment and to preserve linguistic context. Alignment is maintained by noting that the attention curve is predominantly diagonal (in the voice conversion prob... | {
"cite_N": [
"@cite_42",
"@cite_20",
"@cite_16",
"@cite_2"
],
"mid": [
"2950429209",
"2962793481",
"",
"2962834036"
],
"abstract": [
"This paper describes a method based on a sequence-to-sequence learning (Seq2Seq) with attention and context preservation mechanism for voic... |
1907.07769 | 2959758584 | We present a voice conversion solution using recurrent sequence to sequence modeling for DNNs. Our solution takes advantage of recent advances in attention based modeling in the fields of Neural Machine Translation (NMT), Text-to-Speech (TTS) and Automatic Speech Recognition (ASR). The problem consists of converting be... | Developments in the generative modeling (primarily, Variational Autoencoders @cite_5 and Generative Adversarial Networks @cite_21 ) front have led to their use in voice conversion problems. In @cite_11 , a learned similarity metric obtained through a GAN discriminator is used to correct oversmoothed speech that results... | {
"cite_N": [
"@cite_35",
"@cite_21",
"@cite_39",
"@cite_19",
"@cite_5",
"@cite_31",
"@cite_51",
"@cite_20",
"@cite_11"
],
"mid": [
"2202109488",
"1710476689",
"2805669069",
"2608338293",
"",
"",
"2774848319",
"2962793481",
"2747744257"
],
... |
1907.07769 | 2959758584 | We present a voice conversion solution using recurrent sequence to sequence modeling for DNNs. Our solution takes advantage of recent advances in attention based modeling in the fields of Neural Machine Translation (NMT), Text-to-Speech (TTS) and Automatic Speech Recognition (ASR). The problem consists of converting be... | Our work is influenced by recent TTS works involving transfer learning and speaker adaptation. The recently published work @cite_49 demonstrates a methodology to use adapt a trained network for new speakers with a wavenet. Likewise, in @cite_8 , a speaker embedding is extracted using a discriminative network for unseen... | {
"cite_N": [
"@cite_15",
"@cite_49",
"@cite_8"
],
"mid": [
"2788357188",
"2963912924",
"2963432880"
],
"abstract": [
"Voice cloning is a highly desired feature for personalized speech interfaces. Neural network based speech synthesis has been shown to generate high quality speech ... |
1907.08038 | 2956579178 | We propose a novel algorithm to ensure @math -differential privacy for answering range queries on trajectory data. In order to guarantee privacy, differential privacy mechanisms add noise to either data or query, thus introducing errors to queries made and potentially decreasing the utility of information. In contrast ... | @cite_20 define the dependency between cells instead of points by mapping the trajectories to a grid to count the movement frequencies between the adjacent cells. However, a frequency vector only maintains the number of transitions for a group of observations without information about the spatial adjacency of two vecto... | {
"cite_N": [
"@cite_20"
],
"mid": [
"47444322"
],
"abstract": [
"We propose a novel approach to privacy-preserving analytical processing within a distributed setting, and tackle the problem of obtaining aggregated information about vehicle traffic in a city from movement data collected by individ... |
1907.08038 | 2956579178 | We propose a novel algorithm to ensure @math -differential privacy for answering range queries on trajectory data. In order to guarantee privacy, differential privacy mechanisms add noise to either data or query, thus introducing errors to queries made and potentially decreasing the utility of information. In contrast ... | Recently, @cite_13 developed a mechanism named Private Spatial Histogram for range queries on trajectories. publishes a synthetic spatial histogram under @math -differential privacy. It is a query-aware mechanism that extends the idea of in @cite_3 . , takes a spatial histogram and a query set as input and utilizes the... | {
"cite_N": [
"@cite_13",
"@cite_3"
],
"mid": [
"2859661414",
"2170204206"
],
"abstract": [
"Studying trajectories of individuals has received growing interest. The aggregated movement behaviour of people provides important insights about their habits, interests, and lifestyles. Understand... |
1907.07723 | 2959678280 | We study the problem of repeated play in a zero-sum game in which the payoff matrix may change, in a possibly adversarial fashion, on each round; we call these Online Matrix Games. Finding the Nash Equilibrium (NE) of a two player zero-sum game is core to many problems in statistics, optimization, and economics, and fo... | The reader familiar with Online Convex Optimization (OCO) may find it closely related to the OMG problem. In the OCO setting, a player is given a convex, closed, and bounded action set @math , and must repeatedly choose an action @math before the convex function @math is revealed. The player's goal is to obtain subline... | {
"cite_N": [
"@cite_53",
"@cite_22",
"@cite_12",
"@cite_6"
],
"mid": [
"1508384000",
"2118075434",
"2148825261",
"2076798318"
],
"abstract": [
"We introduce an efficient algorithm for the problem of online linear optimization in the bandit setting which achieves the optima... |
1907.07723 | 2959678280 | We study the problem of repeated play in a zero-sum game in which the payoff matrix may change, in a possibly adversarial fashion, on each round; we call these Online Matrix Games. Finding the Nash Equilibrium (NE) of a two player zero-sum game is core to many problems in statistics, optimization, and economics, and fo... | Related to the OMG problem with bandit feedback is the seminal work of @cite_4 . They provide the first sublinear regret bound for Online Convex Optimization with bandit feedback, using a one-point estimate of the gradient. The one-point gradient estimate used in @cite_4 is similar to those independently proposed in @c... | {
"cite_N": [
"@cite_35",
"@cite_4",
"@cite_8",
"@cite_53",
"@cite_39",
"@cite_44",
"@cite_16"
],
"mid": [
"",
"2004001705",
"2116067849",
"1508384000",
"2301614296",
"2012117977",
"2473549844"
],
"abstract": [
"",
"We study a general online conv... |
1901.06026 | 2909735549 | In crowd counting datasets, people appear at different scales, depending on their distance to the camera. To address this issue, we propose a novel multi-branch scale-aware attention network that exploits the hierarchical structure of convolutional neural networks and generates, in a single forward pass, multi-scale de... | Attention models have been widely used for many computer vision tasks like image classification @cite_16 @cite_47 , object detection @cite_19 @cite_20 , semantic segmentation @cite_5 @cite_10 , saliency detection @cite_13 and, very recently, crowd counting @cite_30 . These models work by learning an intermediate attent... | {
"cite_N": [
"@cite_13",
"@cite_30",
"@cite_19",
"@cite_5",
"@cite_47",
"@cite_16",
"@cite_10",
"@cite_20"
],
"mid": [
"2796255296",
"2804654955",
"",
"2158865742",
"2221625691",
"",
"",
"1689909837"
],
"abstract": [
"Contexts play an import... |
1901.06024 | 2950571912 | Benchmarks of bugs are essential to empirically evaluate automatic program repair tools. In this paper, we present Bears, a project for collecting and storing bugs into an extensible bug benchmark for automatic repair studies in Java. The collection of bugs relies on commit building state from Continuous Integration (C... | Benchmarks of bugs are assets that have been used in software bug-related research fields to support empirical evaluations. Several benchmarks were first created for the software testing research community, such as Siemens @cite_18 and SIR @cite_7 , two notable and well-cited benchmarks. The majority of bugs in these t... | {
"cite_N": [
"@cite_18",
"@cite_7"
],
"mid": [
"2134691366",
"1971137495"
],
"abstract": [
"This paper reports an experimental study investigating the effectiveness of two code-based test adequacy criteria for identifying sets of test cases that detect faults. The all-edges and all-DUs (m... |
1901.06024 | 2950571912 | Benchmarks of bugs are essential to empirically evaluate automatic program repair tools. In this paper, we present Bears, a project for collecting and storing bugs into an extensible bug benchmark for automatic repair studies in Java. The collection of bugs relies on commit building state from Continuous Integration (C... | To the best of our knowledge, the first benchmarks proposed for automatic program repair research are ManyBugs and IntroClass @cite_4 . ManyBugs contains 185 bugs collected from nine large, popular, open-source programs. On the other hand, IntroClass targets small programs written by novices, and contains 998 bugs coll... | {
"cite_N": [
"@cite_4"
],
"mid": [
"841012168"
],
"abstract": [
"The field of automated software repair lacks a set of common benchmark problems. Although benchmark sets are used widely throughout computer science, existing benchmarks are not easily adapted to the problem of automatic defect repa... |
1901.06024 | 2950571912 | Benchmarks of bugs are essential to empirically evaluate automatic program repair tools. In this paper, we present Bears, a project for collecting and storing bugs into an extensible bug benchmark for automatic repair studies in Java. The collection of bugs relies on commit building state from Continuous Integration (C... | More recently other benchmarks were proposed for automatic program repair. Codeflaws @cite_5 contains 3,902 bugs extracted from programming contests available on Codeforces. Codeflaws is also for the C language, and the programs range from one to 322 lines of code. QuixBugs @cite_15 is a multi-lingual benchmark, which ... | {
"cite_N": [
"@cite_5",
"@cite_15"
],
"mid": [
"2620986014",
"2762550985"
],
"abstract": [
"Several automated program repair techniques have been proposed to reduce the time and effort spent in bug-fixing. While these repair tools are designed to be generic such that they could address ma... |
1901.06024 | 2950571912 | Benchmarks of bugs are essential to empirically evaluate automatic program repair tools. In this paper, we present Bears, a project for collecting and storing bugs into an extensible bug benchmark for automatic repair studies in Java. The collection of bugs relies on commit building state from Continuous Integration (C... | The closest benchmarks to are Defects4J @cite_12 and Bugs.jar @cite_0 , both for Java. Defects4J contains 395 reproducible bugs collected from six projects, and Bugs.jar contains 1,158 reproducible bugs collected from eight Apache projects. To collect bugs, the approach used for both benchmarks is based on bug tracking... | {
"cite_N": [
"@cite_0",
"@cite_12"
],
"mid": [
"2883977877",
"2156723666"
],
"abstract": [
"We present Bugs.jar, a large-scale dataset for research in automated debugging, patching, and testing of Java programs. Bugs.jar is comprised of 1,158 bugs and patches, drawn from 8 large, popular ... |
1901.06144 | 2910986363 | Accurate, nontrivial quantum operations on many qubits are experimentally challenging. As opposed to the standard approach of compiling larger unitaries into sequences of 2-qubit gates, we propose a protocol on Hamiltonian control fields which implements highly selective multi-qubit gates in a strongly-coupled many-bod... | We previously described a very similar resonantly driven gate in Ref. @cite_12 , which was based on the so-called Krawtchouk spin chain. In the present work, we generalize many aspects of this first result, and show how the same line of reasoning applies to a very different system featuring long-range rather than just ... | {
"cite_N": [
"@cite_12"
],
"mid": [
"2736372133"
],
"abstract": [
"textabstractWe propose a strategy for engineering multiqubit quantum gates. As a first step, it employs an eigengate to map states in the computational basis to eigenstates of a suitable many-body Hamiltonian. The second step empl... |
1901.06144 | 2910986363 | Accurate, nontrivial quantum operations on many qubits are experimentally challenging. As opposed to the standard approach of compiling larger unitaries into sequences of 2-qubit gates, we propose a protocol on Hamiltonian control fields which implements highly selective multi-qubit gates in a strongly-coupled many-bod... | The most obvious competitor of our protocol is conventional compiling of any quantum operation into a universal set of single- and two-qubit gates. Extensive research efforts have greatly optmized compiling methods, and in the asymptotics of many qubits, compiling approach becomes increasingly favorable compared to our... | {
"cite_N": [
"@cite_27"
],
"mid": [
"2755920271"
],
"abstract": [
"To enable a quantum computer to solve practical problems more efficiently than classical computers, quantum programming languages and compilers are required to translate quantum algorithms into machine code; here the currently ava... |
1901.05997 | 2910606695 | Anecdotal evidence has emerged suggesting that state-sponsored organizations, like the Russian Internet Research Agency, have exploited mainstream social. Their primary goal is apparently to conduct information warfare operations to manipulate public opinion using accounts disguised as "normal" people. To increase enga... | Other work has studied state-sponsored accounts' behavior on, and use of, social networks. Specifically, @cite_16 analyze the advertisements purchased by Russian accounts on Facebook. By performing clustering and semantic analysis, they identify their targeted campaigns over time, concluding that their main goal is to ... | {
"cite_N": [
"@cite_0",
"@cite_28",
"@cite_46",
"@cite_16"
],
"mid": [
"2900289476",
"2897648888",
"2786091114",
"2963635075"
],
"abstract": [
"Over the past few years, extensive anecdotal evidence emerged that suggests the involvement of state-sponsored actors (or \"troll... |
1901.05997 | 2910606695 | Anecdotal evidence has emerged suggesting that state-sponsored organizations, like the Russian Internet Research Agency, have exploited mainstream social. Their primary goal is apparently to conduct information warfare operations to manipulate public opinion using accounts disguised as "normal" people. To increase enga... | Finally, @cite_18 use machine learning to detect Twitter users that are likely to share content that originates from Russian state-sponsored accounts. | {
"cite_N": [
"@cite_18"
],
"mid": [
"2950434393"
],
"abstract": [
"Social media, once hailed as a vehicle for democratization and the promotion of positive social change across the globe, are under attack for becoming a tool of political manipulation and spread of disinformation. A case in point ... |
1901.06033 | 2910986402 | The Variational Auto-Encoder (VAE) model is a popular method to learn at once a generative model and embeddings for data living in a high-dimensional space. In the real world, many datasets may be assumed to be hierarchically structured. Traditionally, VAE uses a Euclidean latent space, but tree-like structures cannot ... | In the BNP 's literature, explicitly modelling the hierarchical structure of data has been a long-going trend . Embedding graphs in hyperbolic spaces has been empirically shown to yield a more compact representation compared to Euclidean space, especially for low dimensions. @cite_6 studied the trade-offs of tree embed... | {
"cite_N": [
"@cite_6"
],
"mid": [
"2797520557"
],
"abstract": [
"Hyperbolic embeddings offer excellent quality with few dimensions when embedding hierarchical data structures like synonym or type hierarchies. Given a tree, we give a combinatorial construction that embeds the tree in hyperbolic s... |
1901.06081 | 2911064732 | Abstract This paper presents a novel iterative deep learning framework and applies it to document enhancement and binarization. Unlike the traditional methods that predict the binary label of each pixel on the input image, we train the neural network to learn the degradations in document images and produce uniform imag... | Binarization is a classical research problem for document analysis and many document binarization methods have been proposed over the past two decades in the literature. It aims to convert each pixel in a document image into either text or background. The most popular and simple method is the Otsu @cite_53 , which is a... | {
"cite_N": [
"@cite_30",
"@cite_53",
"@cite_29",
"@cite_39",
"@cite_40",
"@cite_31"
],
"mid": [
"1600276254",
"2133059825",
"2053896220",
"2128060444",
"2072610689",
"2067887288"
],
"abstract": [
"A new and distinct spur type apple variety which originated ... |
1901.06081 | 2911064732 | Abstract This paper presents a novel iterative deep learning framework and applies it to document enhancement and binarization. Unlike the traditional methods that predict the binary label of each pixel on the input image, we train the neural network to learn the degradations in document images and produce uniform imag... | Other priori knowledge of text is also exploit for binarization, such as the edge pixels extracted by edge detectors. For example, the Canny edge detector is used to extract edge pixels in @cite_6 and then the closed image edges are considered as seeds to find the text region. The transition pixel which is a generation... | {
"cite_N": [
"@cite_28",
"@cite_44",
"@cite_24",
"@cite_6"
],
"mid": [
"2013420608",
"2759766068",
"2091863778",
"2412298086"
],
"abstract": [
"This paper introduces a novel binarization method based on the concept of transition pixel, a generalization of edge pixels. Such... |
1901.06081 | 2911064732 | Abstract This paper presents a novel iterative deep learning framework and applies it to document enhancement and binarization. Unlike the traditional methods that predict the binary label of each pixel on the input image, we train the neural network to learn the degradations in document images and produce uniform imag... | Convolutional neural networks achieve good performance on various applications, which is also applied in document analysis. For example, the winner of the recent DIBCO event @cite_23 uses the U-Net convolutional network architecture for accurate pixel classification. @cite_20 , the fully convolutional neural network is... | {
"cite_N": [
"@cite_18",
"@cite_36",
"@cite_23",
"@cite_16",
"@cite_10",
"@cite_20"
],
"mid": [
"2787190313",
"2731900166",
"2785874433",
"2810569655",
"2751352153",
"2964010994"
],
"abstract": [
"Document image binarization is one of the critical initial s... |
1901.06237 | 2909579777 | Due to concerns about human error in crowdsourcing, it is standard practice to collect labels for the same data point from multiple internet workers. We here show that the resulting budget can be used more effectively with a flexible worker assignment strategy that asks fewer workers to analyze easy-to-label data and m... | Related Crowdsourcing Methodologies. Balancing the demands that accuracy requirements and budget limits place on crowdsourcing experiments has been the focus of research in various communities, including machine learning , human computation , data management , and computer vision . The crowdsourcing mechanisms used in ... | {
"cite_N": [
"@cite_0",
"@cite_10",
"@cite_2"
],
"mid": [
"2072514260",
"2963573141",
"2610525153"
],
"abstract": [
"An increasing number of studies in political communication focus on the “sentiment” or “tone” of news content, political speeches, or advertisements. This growing i... |
1901.06237 | 2909579777 | Due to concerns about human error in crowdsourcing, it is standard practice to collect labels for the same data point from multiple internet workers. We here show that the resulting budget can be used more effectively with a flexible worker assignment strategy that asks fewer workers to analyze easy-to-label data and m... | Our work is different from previously-proposed crowdsourcing methodologies with adaptive worker assignments because these assume that the same workers can be employed with user profile tracking.'' The worker-task allocation scheme by @cite_7 relies on being able to incrementally estimate [the workers' accuracy] based o... | {
"cite_N": [
"@cite_3",
"@cite_7",
"@cite_8"
],
"mid": [
"2149906572",
"2096848877",
"2129345386"
],
"abstract": [
"In this paper we address the problem of budget allocation for redundantly crowdsourcing a set of classification tasks where a key challenge is to find a trade-off be... |
1901.06237 | 2909579777 | Due to concerns about human error in crowdsourcing, it is standard practice to collect labels for the same data point from multiple internet workers. We here show that the resulting budget can be used more effectively with a flexible worker assignment strategy that asks fewer workers to analyze easy-to-label data and m... | Our work is distinct from prior work in that our system not only learns an optimal crowd worker allocation that is adapted to task difficulty, but also a mapping from data features to crowd worker allocations. In their award-winning paper, @cite_2 addressed a related data-focused problem -- how to solicit fewer human r... | {
"cite_N": [
"@cite_2"
],
"mid": [
"2610525153"
],
"abstract": [
"Visual question answering systems empower users to ask any question about any image and receive a valid answer. However, existing systems do not yet account for the fact that a visual question can lead to a single answer or multipl... |
1901.06237 | 2909579777 | Due to concerns about human error in crowdsourcing, it is standard practice to collect labels for the same data point from multiple internet workers. We here show that the resulting budget can be used more effectively with a flexible worker assignment strategy that asks fewer workers to analyze easy-to-label data and m... | A flexible crowdsourcing scheme that collects additional labels for tweets that are estimated to be difficult to understand because they contain sarcasm has been proposed by @cite_0 . Their estimation is based on a Natural Language Processing (NLP) analysis, for example, whether the tweet included texting lingo, such a... | {
"cite_N": [
"@cite_0"
],
"mid": [
"2072514260"
],
"abstract": [
"An increasing number of studies in political communication focus on the “sentiment” or “tone” of news content, political speeches, or advertisements. This growing interest in measuring sentiment coincides with a dramatic increase i... |
1901.06237 | 2909579777 | Due to concerns about human error in crowdsourcing, it is standard practice to collect labels for the same data point from multiple internet workers. We here show that the resulting budget can be used more effectively with a flexible worker assignment strategy that asks fewer workers to analyze easy-to-label data and m... | Related Methods for Image Segmentation. Many solutions have been proposed for crowdsourcing the task of image segmentation. The most common proposed solution requires task requesters to collect redundant data from multiple crowd workers and uses majority voting (e.g., majority of the decisions of 5 workers per task @ci... | {
"cite_N": [
"@cite_5",
"@cite_4"
],
"mid": [
"1999045972",
"2055302526"
],
"abstract": [
"Analyses of biomedical images often rely on demarcating the boundaries of biological structures (segmentation). While numerous approaches are adopted to address the segmentation problem including co... |
1901.06199 | 2909896778 | Generative Adversarial Networks (GAN) receive great attentions recently due to its excellent performance in image generation, transformation, and super-resolution. However, GAN has rarely been studied and trained for classification, leading that the generated images may not be appropriate for classification. In this pa... | The research of image super-resolution can be divided into two categories: one is based on single image super-resolution (SISR), and the other is based on multiple image super-resolution (MISR) @cite_14 . Our work can be cast into the first category. We will focus on single image super-resolution (SISR) and will not fu... | {
"cite_N": [
"@cite_14"
],
"mid": [
"1811400895"
],
"abstract": [
"Growing interest in super-resolution (SR) restoration of video sequences and the closed related problem of construction of SR still images from image sequences has led to the emergence of several competing methodologies. We review... |
1901.06199 | 2909896778 | Generative Adversarial Networks (GAN) receive great attentions recently due to its excellent performance in image generation, transformation, and super-resolution. However, GAN has rarely been studied and trained for classification, leading that the generated images may not be appropriate for classification. In this pa... | Recently, convolutional neural network (CNN) based SR algorithms have shown excellent performance. In @cite_1 , the authors encoded a sparse representation prior into a feed-forward network architecture based on the learned iterative shrinkage and thresholding algorithm (LISTA) @cite_8 . @cite_24 @cite_7 used bicubic i... | {
"cite_N": [
"@cite_7",
"@cite_8",
"@cite_1",
"@cite_24",
"@cite_10",
"@cite_25",
"@cite_17"
],
"mid": [
"",
"2118103795",
"1919542679",
"54257720",
"2331128040",
"2196707239",
"2214802144"
],
"abstract": [
"",
"In Sparse Coding (SC), input vect... |
1901.06199 | 2909896778 | Generative Adversarial Networks (GAN) receive great attentions recently due to its excellent performance in image generation, transformation, and super-resolution. However, GAN has rarely been studied and trained for classification, leading that the generated images may not be appropriate for classification. In this pa... | Generative Adversarial Nets (GAN) is proposed by Goodfellow @cite_6 which contains two parts, a generator and a discriminator. The generator is responsible for generating images close to the real pictures to fool the discriminator, and the discriminator is responsible to discriminate the picture from the generator or r... | {
"cite_N": [
"@cite_28",
"@cite_6"
],
"mid": [
"2963341057",
"2099471712"
],
"abstract": [
"Adversarial examples are augmented data points generated by imperceptible perturbation of input samples. They have recently drawn much attention with the machine learning and data mining community.... |
1901.06199 | 2909896778 | Generative Adversarial Networks (GAN) receive great attentions recently due to its excellent performance in image generation, transformation, and super-resolution. However, GAN has rarely been studied and trained for classification, leading that the generated images may not be appropriate for classification. In this pa... | In 2016, @cite_21 proposed DCGAN which is stable in most settings and shows the vector arithmetics as an intrinsic property of the representations learned by the Generator. @cite_23 proposed the conditional GAN, the idea is to use labels for some data to help network build salient representations, it can control the ge... | {
"cite_N": [
"@cite_21",
"@cite_0",
"@cite_19",
"@cite_27",
"@cite_23"
],
"mid": [
"2173520492",
"2964218010",
"",
"2523714292",
"2125389028"
],
"abstract": [
"In recent years, supervised learning with convolutional networks (CNNs) has seen huge adoption in compute... |
1907.07469 | 2960922699 | In this work, we propose an edge detection algorithm by estimating a lifetime of an event produced from dynamic vision sensor (DVS), also known as event camera. The event camera, unlike traditional CMOS camera, generates sparse event data at a pixel whose log-intensity changes. Due to this characteristic, theoretically... | Some research aim to detect edges, not just line segments that are frequently found in artifacts. F. Barranco al @cite_0 detects the contour of foreground objects. They extract features from the accumulated events such as orientation, timestamp, motion, and time texture. Then the boundary is predicted from the learned ... | {
"cite_N": [
"@cite_0",
"@cite_4",
"@cite_8"
],
"mid": [
"2216124221",
"1975697167",
"1539361405"
],
"abstract": [
"The bio-inspired, asynchronous event-based dynamic vision sensor records temporal changes in the luminance of the scene at high temporal resolution. Since events are... |
1907.07581 | 2958911020 | Online personalized news product needs a suitable cover for the article. The news cover demands to be with high image quality, and draw readers' attention at same time, which is extraordinary challenging due to the subjectivity of the task. In this paper, we assess the news cover from image clarity and object salience ... | . Human visual system is highly sensitive to edge and contour information of an image @cite_17 . Some IQA studies take edge structure information as the main image quality consideration, for example, in @cite_13 the authors apply edge information for both blur and noise detection, which are the major factors on image q... | {
"cite_N": [
"@cite_19",
"@cite_13",
"@cite_17"
],
"mid": [
"2508724573",
"",
"2032586563"
],
"abstract": [
"Since the human visual system (HVS) is highly sensitive to edges, a novel image quality assessment (IQA) metric for assessing screen content images (SCIs) is proposed in th... |
1907.07581 | 2958911020 | Online personalized news product needs a suitable cover for the article. The news cover demands to be with high image quality, and draw readers' attention at same time, which is extraordinary challenging due to the subjectivity of the task. In this paper, we assess the news cover from image clarity and object salience ... | In recent years, the idea of employing a CNN based approach for no-reference IQA (NR-IQA) tasks is arising, and meanwhile the performance of NR-IQA has been significantly improved under such methods @cite_24 @cite_14 . For example, in @cite_14 , a CNN is directly utilized for image quality prediction without a referenc... | {
"cite_N": [
"@cite_14",
"@cite_22",
"@cite_7",
"@cite_8",
"@cite_24",
"@cite_20"
],
"mid": [
"2051596736",
"2964309882",
"2194775991",
"2531409750",
"2894561647",
"2754213847"
],
"abstract": [
"In this work we describe a Convolutional Neural Network (CNN) ... |
1907.07581 | 2958911020 | Online personalized news product needs a suitable cover for the article. The news cover demands to be with high image quality, and draw readers' attention at same time, which is extraordinary challenging due to the subjectivity of the task. In this paper, we assess the news cover from image clarity and object salience ... | . MTL is based on a fundamental idea that different tasks could share a common low level representation. In many computer vision tasks, MTL has exhibited advantages in performance improvement and memory saving. In @cite_6 , one unified architecture which jointly learn low-, mid-, and high-level vision tasks is introduc... | {
"cite_N": [
"@cite_21",
"@cite_3",
"@cite_6"
],
"mid": [
"2963877604",
"125693051",
"2963498646"
],
"abstract": [
"Multi-task learning in Convolutional Networks has displayed remarkable success in the field of recognition. This success can be largely attributed to learning shared... |
1907.07671 | 2958881886 | Stress research is a rapidly emerging area in thefield of electroencephalography (EEG) based signal processing.The use of EEG as an objective measure for cost effective andpersonalized stress management becomes important in particularsituations such as the non-availability of mental health this http URL this study, lon... | Hemispheric specialization is a major concern in neuro-physiological research. Generally, a healthy brain at rest has a fairly balanced level of activity in both hemispheres of brain @cite_30 . The left hemisphere is associated with the processing of positive emotions, while the right hemisphere is associated with the ... | {
"cite_N": [
"@cite_30",
"@cite_32",
"@cite_31",
"@cite_16",
"@cite_12"
],
"mid": [
"64229777",
"1997785567",
"1495009241",
"2133085034",
"2040325003"
],
"abstract": [
"",
"Abstract Spontaneously occurring brief periods of lower voltage irregular activity occur... |
1907.07543 | 2960456850 | Despite the recent success of deep transfer learning approaches in NLP, there is a lack of quantitative studies demonstrating the gains these models offer in low-shot text classification tasks over existing paradigms. Deep transfer learning approaches such as BERT and ULMFiT demonstrate that they can beat state-of-the-... | It is well established that there is no single classical machine learning classifier that consistently achieves the best classification performance. For example between the works in @cite_17 @cite_4 @cite_3 they showed that various classical machine learning approaches all slightly out performed each other. This is a l... | {
"cite_N": [
"@cite_4",
"@cite_3",
"@cite_17"
],
"mid": [
"1823790170",
"",
"2595653137"
],
"abstract": [
"Cyberbullying is becoming a major concern in online environments with troubling consequences. However, most of the technical studies have focused on the detection of cyberbul... |
1907.07613 | 2960281739 | Template-matching methods for visual tracking have gained popularity recently due to their good performance and fast speed. However, they lack effective ways to adapt to changes in the target object's appearance, making their tracking accuracy still far from state-of-the-art. In this paper, we propose a dynamic memory ... | In this section, we review related work on tracking-by-detection, tracking by template-matching, memory networks and multi-task learning. A preliminary version of our work appears in ECCV 2018 @cite_63 . This paper contains additional improvements in both methodology and experiments, including: 1) we propose a negative... | {
"cite_N": [
"@cite_63"
],
"mid": [
"2963471260"
],
"abstract": [
"Template-matching methods for visual tracking have gained popularity recently due to their comparable performance and fast speed. However, they lack effective ways to adapt to changes in the target object’s appearance, making thei... |
1907.07613 | 2960281739 | Template-matching methods for visual tracking have gained popularity recently due to their good performance and fast speed. However, they lack effective ways to adapt to changes in the target object's appearance, making their tracking accuracy still far from state-of-the-art. In this paper, we propose a dynamic memory ... | Tracking-by-detection treats object tracking as a detection problem within an ROI image, where an online learned classifier is used to distinguish the target from the background. The difficulty of updating the classifier to adapt to appearance variations is that the bounding box predicted on each frame may not be accur... | {
"cite_N": [
"@cite_80",
"@cite_23",
"@cite_50"
],
"mid": [
"2124211486",
"2109579504",
"1807914171"
],
"abstract": [
"This paper investigates long-term tracking of unknown objects in a video stream. The object is defined by its location and extent in a single frame. In every fram... |
1907.07613 | 2960281739 | Template-matching methods for visual tracking have gained popularity recently due to their good performance and fast speed. However, they lack effective ways to adapt to changes in the target object's appearance, making their tracking accuracy still far from state-of-the-art. In this paper, we propose a dynamic memory ... | With the widespread use of CNNs in the computer vision community, many methods @cite_1 have applied CNNs as the classifier to localize the target. @cite_49 uses two fully convolutional neural networks to estimate the target's bounding box, including a GNet that captures category information and an SNet that classifies ... | {
"cite_N": [
"@cite_29",
"@cite_1",
"@cite_40",
"@cite_49",
"@cite_12"
],
"mid": [
"1857884451",
"2767302379",
"2737572441",
"2211629196",
"2095705004"
],
"abstract": [
"We propose a novel visual tracking algorithm based on the representations from a discriminative... |
1907.07613 | 2960281739 | Template-matching methods for visual tracking have gained popularity recently due to their good performance and fast speed. However, they lack effective ways to adapt to changes in the target object's appearance, making their tracking accuracy still far from state-of-the-art. In this paper, we propose a dynamic memory ... | Matching-based methods have recently gained popularity due to their fast speed and promising performance. The most notable is the fully convolutional Siamese network (SiamFC) @cite_58 . Although it only uses the first frame as the template, SiamFC achieves competitive results and fast speed. The key deficiency of SiamF... | {
"cite_N": [
"@cite_30",
"@cite_18",
"@cite_64",
"@cite_58",
"@cite_10"
],
"mid": [
"2962824803",
"2963873961",
"2963854930",
"2470394683",
"2776035257"
],
"abstract": [
"The Correlation Filter is an algorithm that trains a linear template to discriminate between i... |
1907.07613 | 2960281739 | Template-matching methods for visual tracking have gained popularity recently due to their good performance and fast speed. However, they lack effective ways to adapt to changes in the target object's appearance, making their tracking accuracy still far from state-of-the-art. In this paper, we propose a dynamic memory ... | To further improve the speed of SiamFC, @cite_55 reduces the feature computation cost for easy frames, by using deep reinforcement learning to train policies for early stopping the feed-forward calculations of the CNN when the response confidence is high enough. SINT @cite_43 also uses Siamese networks for visual track... | {
"cite_N": [
"@cite_43",
"@cite_55",
"@cite_65",
"@cite_15"
],
"mid": [
"2408241409",
"2963791342",
"2964253307",
"2562852216"
],
"abstract": [
"In this paper we present a tracker, which is radically different from state-of-the-art trackers: we apply no model updating, no ... |
1907.07613 | 2960281739 | Template-matching methods for visual tracking have gained popularity recently due to their good performance and fast speed. However, they lack effective ways to adapt to changes in the target object's appearance, making their tracking accuracy still far from state-of-the-art. In this paper, we propose a dynamic memory ... | Multi-task learning has been successfully used in many applications of machine learning, ranging from natural language processing @cite_37 and speech recognition @cite_81 to computer vision @cite_39 . @cite_70 estimates the street direction in an autonomous driving car by predicting various characteristics of the road,... | {
"cite_N": [
"@cite_37",
"@cite_7",
"@cite_70",
"@cite_53",
"@cite_39",
"@cite_44",
"@cite_77",
"@cite_81",
"@cite_45",
"@cite_83"
],
"mid": [
"2117130368",
"2052678124",
"",
"",
"",
"1648933886",
"",
"2091432990",
"2137881638",
"190... |
1907.07647 | 2960675232 | Particle Swarm Optimisation (PSO) is a powerful optimisation algorithm that can be used to locate global maxima in a search space. Recent interest in swarms of Micro Aerial Vehicles (MAVs) begs the question as to whether PSO can be used as a method to enable real robotic swarms to locate a target goal point. However, t... | We begin by reviewing the most related work in the area of PSO, Potential Field methods and Flocking. PSO itself is a vast field with applications in many different areas (see @cite_22 for details), our aim here is not to cover the entirety of this but only what is relevant to aerial and swarm robotics. We also review ... | {
"cite_N": [
"@cite_22"
],
"mid": [
"1859314164"
],
"abstract": [
"Particle swarm optimization (PSO) is a heuristic global optimization method, proposed originally by Kennedy and Eberhart in 1995. It is now one of the most commonly used optimization techniques. This survey presented a comprehensi... |
1907.07647 | 2960675232 | Particle Swarm Optimisation (PSO) is a powerful optimisation algorithm that can be used to locate global maxima in a search space. Recent interest in swarms of Micro Aerial Vehicles (MAVs) begs the question as to whether PSO can be used as a method to enable real robotic swarms to locate a target goal point. However, t... | PSO has been applied to Unmanned Aerial Vehicles (UAVs) and MAVs in various ways already. Optimal route planning for MAVs is an optimisation problem that is tackled in @cite_19 @cite_1 by constructing complex fitness functions consisting of a number of different metrics that would affect the success of an MAV carrying ... | {
"cite_N": [
"@cite_9",
"@cite_1",
"@cite_6",
"@cite_19",
"@cite_16"
],
"mid": [
"836179420",
"",
"2471940543",
"2053726668",
"2023086460"
],
"abstract": [
"This work presents a method to build a robust controller for a hose transportation system performed by aeria... |
1907.07647 | 2960675232 | Particle Swarm Optimisation (PSO) is a powerful optimisation algorithm that can be used to locate global maxima in a search space. Recent interest in swarms of Micro Aerial Vehicles (MAVs) begs the question as to whether PSO can be used as a method to enable real robotic swarms to locate a target goal point. However, t... | The work most related to ours in theoretical approach is @cite_18 . In this work each individual ePuck robot represents a particle in the PSO algorithm where the aim is to find an area of interest. However, the main contributions of our work compared to @cite_18 is that we extend this model to 3 dimensions for aerial v... | {
"cite_N": [
"@cite_18",
"@cite_7"
],
"mid": [
"2112571153",
"2783023810"
],
"abstract": [
"Within the field of multi-robot systems, multi-robot search is one area which is currently receiving a lot of research attention. One major challenge within this area is to design effective algorit... |
1907.07377 | 2959120033 | A Controller Area Network (CAN) bus in the vehicles is an efficient standard bus enabling communication between all Electronic Control Units (ECU). However, CAN bus is not enough to protect itself because of lack of security features. To detect suspicious network connections effectively, the intrusion detection system ... | The early research for anomaly detection of the in-vehicle system was introduced by Hoppe @cite_3 . He presented three selected characteristics as patterns available for anomaly detection that include the recognition of an increased frequency of cyclic CAN messages, the observation of low-level communication characteri... | {
"cite_N": [
"@cite_0",
"@cite_6",
"@cite_3",
"@cite_2"
],
"mid": [
"2584565207",
"2739928414",
"2148974412",
"2756382106"
],
"abstract": [
"",
"This paper proposes a novel intrusion detection algorithm that aims to identify malicious CAN messages injected by attackers... |
1907.07377 | 2959120033 | A Controller Area Network (CAN) bus in the vehicles is an efficient standard bus enabling communication between all Electronic Control Units (ECU). However, CAN bus is not enough to protect itself because of lack of security features. To detect suspicious network connections effectively, the intrusion detection system ... | Many security research in various fields has adopted deep-learning methods for IDS. For example, Zhang presented a deep-learning method to detect Web attacks by using the specially designed CNN @cite_7 . The method is based on analyzing the HTTP request packets, to which only some preprocessing is needed whereas the te... | {
"cite_N": [
"@cite_4",
"@cite_7"
],
"mid": [
"2599354622",
"2765438277"
],
"abstract": [
"Obtaining models that capture imaging markers relevant for disease progression and treatment monitoring is challenging. Models are typically based on large amounts of data with annotated examples of... |
1907.07202 | 2959373581 | Human gaze is known to be a strong indicator of underlying human intentions and goals during manipulation tasks. This work studies gaze patterns of human teachers demonstrating tasks to robots and proposes ways in which such patterns can be used to enhance robot learning. Using both kinesthetic teaching and video demon... | There is also a rich body of work on eye gaze for human-robot interaction @cite_9 . use nonverbal cues including gaze to study timing coordination between humans and robots. Gaze information has also been shown to enable the establishment of joint attention between the human and robot partner, the recognition of human ... | {
"cite_N": [
"@cite_9"
],
"mid": [
"2617211984"
],
"abstract": [
"This article reviews the state of the art in social eye gaze for human-robot interaction (HRI). It establishes three categories of gaze research in HRI, defined by differences in goals and methods: a human-centered approach, which ... |
1907.07202 | 2959373581 | Human gaze is known to be a strong indicator of underlying human intentions and goals during manipulation tasks. This work studies gaze patterns of human teachers demonstrating tasks to robots and proposes ways in which such patterns can be used to enhance robot learning. Using both kinesthetic teaching and video demon... | There has also been some recent work on utilizing human eye gaze for learning algorithms. used demonstrations from a person wearing an eye tracking hardware along with an egocentric camera to simultaneously ground symbols to their instances in the environment and learn the appearance of such object instances. use gaze ... | {
"cite_N": [
"@cite_8"
],
"mid": [
"1591675293"
],
"abstract": [
"Inverse Reinforcement Learning (IRL) is the problem of learning the reward function underlying a Markov Decision Process given the dynamics of the system and the behaviour of an expert. IRL is motivated by situations where knowledg... |
1907.07384 | 2962029365 | Mutual information has been successfully adopted in filter feature-selection methods to assess both the relevancy of a subset of features in predicting the target variable and the redundancy with respect to other variables. However, existing algorithms are mostly heuristic and do not offer any guarantee on the proposed... | A related theoretical study of feature selection via MI has been recently proposed by @cite_25 . The authors show that the problem of finding the minimal feature subset such that the conditional likelihood of the targets is maximized is equivalent to minimizing the CMI. Based on this result, common heuristics for infor... | {
"cite_N": [
"@cite_25"
],
"mid": [
"2156504490"
],
"abstract": [
"We present a unifying framework for information theoretic feature selection, bringing almost two decades of research on heuristic filter criteria under a single theoretical interpretation. This is in response to the question: \"wh... |
1907.07384 | 2962029365 | Mutual information has been successfully adopted in filter feature-selection methods to assess both the relevancy of a subset of features in predicting the target variable and the redundancy with respect to other variables. However, existing algorithms are mostly heuristic and do not offer any guarantee on the proposed... | In the information theory literature, @cite_13 also analyzes the connection between CMI and minimum mean square error, deriving a similar result to our Theorem . However, classification problems (i.e., minimum zero-one loss) are not considered and the focus is not on feature selection. | {
"cite_N": [
"@cite_13"
],
"mid": [
"2164714023"
],
"abstract": [
"In addition to exploring its various regularity properties, we show that the minimum mean-square error (MMSE) is a concave functional of the input-output joint distribution. In the case of additive Gaussian noise, the MMSE is show... |
1907.07384 | 2962029365 | Mutual information has been successfully adopted in filter feature-selection methods to assess both the relevancy of a subset of features in predicting the target variable and the redundancy with respect to other variables. However, existing algorithms are mostly heuristic and do not offer any guarantee on the proposed... | The authors of @cite_17 propose a nearest neighbor estimator for the CMI and show how it can be used in a classic forward feature selection algorithm. One of the authors' questions is how to devise a suitable stopping condition for such methods. Here we propose a possible answer: our stopping criterion (Section ) is in... | {
"cite_N": [
"@cite_17"
],
"mid": [
"2044956198"
],
"abstract": [
"Mutual information (MI) is used in feature selection to evaluate two key-properties of optimal features, the relevance of a feature to the class variable and the redundancy of similar features. Conditional mutual information (CMI)... |
1907.07384 | 2962029365 | Mutual information has been successfully adopted in filter feature-selection methods to assess both the relevancy of a subset of features in predicting the target variable and the redundancy with respect to other variables. However, existing algorithms are mostly heuristic and do not offer any guarantee on the proposed... | Several existing approaches use linear correlation measures to score the different features @cite_31 @cite_11 @cite_5 @cite_6 @cite_7 . Such algorithms are mostly based on the heuristic intuition that a good feature should be highly correlated with the class and lowly correlated with the other features. Instead, we pro... | {
"cite_N": [
"@cite_7",
"@cite_6",
"@cite_5",
"@cite_31",
"@cite_11"
],
"mid": [
"33196642",
"2013396429",
"1661871015",
"2054033094",
"1495061682"
],
"abstract": [
"Feature selection is a preprocessing phase to machine learning, which leads to increase the classif... |
1907.07240 | 2966730110 | Social media has become an integral part of our daily lives. During time-critical events, the public shares a variety of posts on social media including reports for resource needs, damages, and help offerings for the affected community. Such posts can be relevant and may contain valuable situational awareness informati... | There has been extensive research on the topic of social media for emergency management in the last decade @cite_5 @cite_6 . The nature of data generated over social media has such a high volume, variety, and velocity causing the challenges of Big Crisis Data'' that often overwhelm the emergency services @cite_6 . The ... | {
"cite_N": [
"@cite_17",
"@cite_6",
"@cite_0",
"@cite_5",
"@cite_11"
],
"mid": [
"2472441007",
"2493475353",
"2898274636",
"1934362406",
"2573190752"
],
"abstract": [
"Crisis informatics is a multidisciplinary field combining computing and social science knowledge ... |
1907.07240 | 2966730110 | Social media has become an integral part of our daily lives. During time-critical events, the public shares a variety of posts on social media including reports for resource needs, damages, and help offerings for the affected community. Such posts can be relevant and may contain valuable situational awareness informati... | Among the social media analytics approaches, researchers have modeled public behavior in specific emergencies, addressed the problems of data collection and filtering, classification and summarization as well as visualization of analyzed data for decision support @cite_5 . However, the focus of such works has centered ... | {
"cite_N": [
"@cite_4",
"@cite_5",
"@cite_15",
"@cite_13",
"@cite_12"
],
"mid": [
"2768447691",
"1934362406",
"2962870381",
"2963002801",
"2626583996"
],
"abstract": [
"Rapid access to situation-sensitive data through social media networks creates new opportunities... |
1907.07378 | 2957866194 | Competency Questions (CQs) for an ontology and similar artefacts aim to provide insights into the contents of an ontology and to demarcate its scope. The absence of a controlled natural language, tooling and automation to support the authoring of CQs has hampered their effective use in ontology development and evaluati... | Given that a CNL for CQs is supposed to function for specifying requirements for any ontology, the logic-based knowledge representation must be decoupled from the natural language. At the same time, it is well-known that the other extreme---free-form sentences---makes it exceedingly hard to formalise, be this for query... | {
"cite_N": [
"@cite_7"
],
"mid": [
"2888712326"
],
"abstract": [
"In the context of the development of a virtual tutor to support distance learning courses, this paper presents an approach to solve the problem of automatically answering questions posed by students in a natural language (Portugues... |
1907.07378 | 2957866194 | Competency Questions (CQs) for an ontology and similar artefacts aim to provide insights into the contents of an ontology and to demarcate its scope. The absence of a controlled natural language, tooling and automation to support the authoring of CQs has hampered their effective use in ontology development and evaluati... | CNLs for computation have been proposed as a solution for various information management aspects, such as query formulation to hide SPARQL syntax (e.g., Sparklis @cite_20 and Quelo @cite_15 ), generation of pseudo-NL sentences from axioms in an ontology to formalise them (e.g., ACE @cite_22 ), and software requirements... | {
"cite_N": [
"@cite_22",
"@cite_3",
"@cite_24",
"@cite_19",
"@cite_15",
"@cite_16",
"@cite_20"
],
"mid": [
"1562074255",
"2101493333",
"",
"1832719046",
"2141934106",
"2330889199",
"2260667602"
],
"abstract": [
"This technical report describes the d... |
1907.07171 | 2959108703 | An open secret in contemporary machine learning is that many models work beautifully on standard benchmarks but fail to generalize outside the lab. This has been attributed to training on biased data, which provide poor coverage over real world events. Generative models are no exception, but recent advances in generati... | Biases from training data and network architecture both factor into the generalization capacity of learned models @cite_16 @cite_8 @cite_17 . Dataset biases partly comes from human preferences in taking photos: we typically capture images in specific canonical'' views that are not fully representative of the entire vis... | {
"cite_N": [
"@cite_8",
"@cite_6",
"@cite_16",
"@cite_13",
"@cite_20",
"@cite_17"
],
"mid": [
"2952610664",
"2807007689",
"2031342017",
"1987947320",
"2101734194",
"2957285709"
],
"abstract": [
"Convolutional Neural Networks (CNNs) are commonly thought to r... |
1907.07171 | 2959108703 | An open secret in contemporary machine learning is that many models work beautifully on standard benchmarks but fail to generalize outside the lab. This has been attributed to training on biased data, which provide poor coverage over real world events. Generative models are no exception, but recent advances in generati... | The recent progress in generative models has enabled interesting applications for content creation @cite_11 @cite_14 , including variants that enable end users to control and fine-tune the generated output @cite_3 @cite_19 @cite_12 . A by-product the current work is to further enable users to modify various image prope... | {
"cite_N": [
"@cite_14",
"@cite_3",
"@cite_24",
"@cite_19",
"@cite_12",
"@cite_11"
],
"mid": [
"2904367110",
"",
"2519536754",
"",
"2901107321",
"2952716587"
],
"abstract": [
"We propose an alternative generator architecture for generative adversarial netwo... |
1907.07171 | 2959108703 | An open secret in contemporary machine learning is that many models work beautifully on standard benchmarks but fail to generalize outside the lab. This has been attributed to training on biased data, which provide poor coverage over real world events. Generative models are no exception, but recent advances in generati... | We note a few concurrent papers that also explore trajectories in GAN latent space. @cite_23 learns linear walks in the latent space that correspond to various facial characteristics; they use these walks to measure biases in facial attribute detectors, whereas we study biases in the generative model that originate fro... | {
"cite_N": [
"@cite_25",
"@cite_23",
"@cite_2"
],
"mid": [
"2963577681",
"2950113871",
"2950419363"
],
"abstract": [
"Despite the recent advance of Generative Adversarial Networks (GANs) in high-fidelity image synthesis, there lacks enough understandings on how GANs are able to ma... |
1907.07349 | 2960672606 | Edge computing in the Internet of Things brings applications and content closer to the users by introducing an additional computational layer at the network infrastructure, between cloud and the resource-constrained data producing devices and user equipment. This way, the opportunistic nature of the operational environ... | Previous studies on edge server placement have focused on algorithms for clustering access points with the aim to find candidate locations for the servers as cluster heads. In these works, clustering was based on k-means @cite_24 , k-means with mixed-integer quadratic programming @cite_41 , graph theory as in minimum d... | {
"cite_N": [
"@cite_30",
"@cite_14",
"@cite_41",
"@cite_1",
"@cite_3",
"@cite_24",
"@cite_15",
"@cite_16",
"@cite_11"
],
"mid": [
"2792782202",
"2739769312",
"2916014425",
"2947632778",
"2887279295",
"2597750431",
"2906068868",
"2514431540",
... |
1907.07349 | 2960672606 | Edge computing in the Internet of Things brings applications and content closer to the users by introducing an additional computational layer at the network infrastructure, between cloud and the resource-constrained data producing devices and user equipment. This way, the opportunistic nature of the operational environ... | The computing capacities of edge servers were assumed equal and fixed, except in works @cite_24 @cite_15 , that allowed scaling of the server capacity on-demand to distribute workload evenly, regardless of the resulting cluster size. In @cite_3 , no strict capacity limits were set for servers, but excessive workload ca... | {
"cite_N": [
"@cite_24",
"@cite_15",
"@cite_16",
"@cite_3"
],
"mid": [
"2597750431",
"2906068868",
"2514431540",
"2887279295"
],
"abstract": [
"A device for selectively moistening the flap of envelopes has a moistening member and a pivoting moistening deflector selectively... |
1907.07349 | 2960672606 | Edge computing in the Internet of Things brings applications and content closer to the users by introducing an additional computational layer at the network infrastructure, between cloud and the resource-constrained data producing devices and user equipment. This way, the opportunistic nature of the operational environ... | To maintain the sufficient QoS within the budget limitations, two main approaches were used for determining the required number of servers. First, a tolerated distance from server was decided and the number of servers was minimized given that the distance constraint is met for each access point @cite_15 @cite_30 @cite_... | {
"cite_N": [
"@cite_30",
"@cite_14",
"@cite_24",
"@cite_49",
"@cite_2",
"@cite_15",
"@cite_16",
"@cite_20",
"@cite_11"
],
"mid": [
"2792782202",
"2739769312",
"2597750431",
"",
"2894009982",
"2906068868",
"2514431540",
"",
"2809740924"
],
... |
1907.07349 | 2960672606 | Edge computing in the Internet of Things brings applications and content closer to the users by introducing an additional computational layer at the network infrastructure, between cloud and the resource-constrained data producing devices and user equipment. This way, the opportunistic nature of the operational environ... | Scalability was considered from the algorithmic scalability and the resulting deployment capacity perspectives. First, the algorithmic scalability was exemplified by the number of access points and edge servers. Basic k-means algorithm was applied in @cite_24 without capacity constraints and hierarchical clustering in ... | {
"cite_N": [
"@cite_30",
"@cite_41",
"@cite_1",
"@cite_3",
"@cite_24",
"@cite_16"
],
"mid": [
"2792782202",
"2916014425",
"2947632778",
"2887279295",
"2597750431",
"2514431540"
],
"abstract": [
"Mobile edge computing (MEC) is an emerging technology that aim... |
1907.07349 | 2960672606 | Edge computing in the Internet of Things brings applications and content closer to the users by introducing an additional computational layer at the network infrastructure, between cloud and the resource-constrained data producing devices and user equipment. This way, the opportunistic nature of the operational environ... | If the aim was to minimize the number of servers, optimization was carried out, for example, with different thresholds of distance @cite_15 @cite_45 or capacity @cite_15 @cite_35 . The effects of capacity constraints on intra-cluster traffic and temporal changes on workload balance is investigated in @cite_30 . In @cit... | {
"cite_N": [
"@cite_30",
"@cite_35",
"@cite_45",
"@cite_2",
"@cite_15",
"@cite_16"
],
"mid": [
"2792782202",
"",
"",
"2894009982",
"2906068868",
"2514431540"
],
"abstract": [
"Mobile edge computing (MEC) is an emerging technology that aims at pushing applic... |
1907.07349 | 2960672606 | Edge computing in the Internet of Things brings applications and content closer to the users by introducing an additional computational layer at the network infrastructure, between cloud and the resource-constrained data producing devices and user equipment. This way, the opportunistic nature of the operational environ... | Simulated data sets were utilized in @cite_15 @cite_32 , where the other studies utilized real-world data sets. The data set @cite_47 consists of geo-referenced phone call detail records over the city of Milan for three months' period, which was used in @cite_24 @cite_30 @cite_21 . The Shanghai Telecom data set contain... | {
"cite_N": [
"@cite_30",
"@cite_47",
"@cite_41",
"@cite_36",
"@cite_21",
"@cite_1",
"@cite_32",
"@cite_24",
"@cite_0",
"@cite_15",
"@cite_16",
"@cite_11"
],
"mid": [
"2792782202",
"2190432600",
"2916014425",
"",
"",
"2947632778",
"",
... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.