arxiv_id
float64
1.5k
2.51k
title
stringlengths
9
178
authors
stringlengths
2
22.8k
categories
stringlengths
4
146
summary
stringlengths
103
1.92k
published
stringdate
2015-02-06 10:44:00
2025-07-10 17:59:58
comments
stringlengths
2
417
journal_ref
stringclasses
321 values
doi
stringclasses
398 values
ss_title
stringlengths
8
159
ss_authors
stringlengths
11
8.38k
ss_year
float64
2.02k
2.03k
ss_venue
stringclasses
281 values
ss_citationCount
float64
0
134k
ss_referenceCount
float64
0
429
ss_fieldsOfStudy
stringclasses
47 values
2,108.1125
YOLOP: You Only Look Once for Panoptic Driving Perception
['Dong Wu', 'Manwen Liao', 'Weitian Zhang', 'Xinggang Wang', 'Xiang Bai', 'Wenqing Cheng', 'Wenyu Liu']
['cs.CV']
A panoptic driving perception system is an essential part of autonomous driving. A high-precision and real-time perception system can assist the vehicle in making the reasonable decision while driving. We present a panoptic driving perception network (YOLOP) to perform traffic object detection, drivable area segmentati...
2021-08-25T14:19:42Z
null
[J]. Machine Intelligence Research, 2022: 1-13
10.1007/s11633-022-1339-y
null
null
null
null
null
null
null
2,108.12009
EmoBERTa: Speaker-Aware Emotion Recognition in Conversation with RoBERTa
['Taewoon Kim', 'Piek Vossen']
['cs.CL']
We present EmoBERTa: Speaker-Aware Emotion Recognition in Conversation with RoBERTa, a simple yet expressive scheme of solving the ERC (emotion recognition in conversation) task. By simply prepending speaker names to utterances and inserting separation tokens between the utterances in a dialogue, EmoBERTa can learn int...
2021-08-26T19:34:26Z
4 pages, not including references and appendix
null
null
EmoBERTa: Speaker-Aware Emotion Recognition in Conversation with RoBERTa
['Taewoon Kim', 'Piek Vossen']
2,021
arXiv.org
102
31
['Computer Science']
2,108.12409
Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation
['Ofir Press', 'Noah A. Smith', 'Mike Lewis']
['cs.CL']
Since the introduction of the transformer model by Vaswani et al. (2017), a fundamental question has yet to be answered: how does a model achieve extrapolation at inference time for sequences that are longer than it saw during training? We first show that extrapolation can be enabled by simply changing the position rep...
2021-08-27T17:35:06Z
null
null
null
null
null
null
null
null
null
null
2,108.12626
HeadlineCause: A Dataset of News Headlines for Detecting Causalities
['Ilya Gusev', 'Alexey Tikhonov']
['cs.CL', 'cs.LG']
Detecting implicit causal relations in texts is a task that requires both common sense and world knowledge. Existing datasets are focused either on commonsense causal reasoning or explicit causal relations. In this work, we present HeadlineCause, a dataset for detecting implicit causal relations between pairs of news h...
2021-08-28T11:12:49Z
null
null
null
null
null
null
null
null
null
null
2,108.1296
LOT: A Story-Centric Benchmark for Evaluating Chinese Long Text Understanding and Generation
['Jian Guan', 'Zhuoer Feng', 'Yamei Chen', 'Ruilin He', 'Xiaoxi Mao', 'Changjie Fan', 'Minlie Huang']
['cs.CL']
Standard multi-task benchmarks are essential for developing pretraining models that can generalize to various downstream tasks. Existing benchmarks for natural language processing (NLP) usually focus only on understanding or generating short texts. However, long text modeling requires many distinct abilities in contras...
2021-08-30T02:38:32Z
Accepted by TACL 2022. Benchmark datasets, pretraining models, appendix url: https://github.com/thu-coai/LOT-LongLM
null
null
null
null
null
null
null
null
null
2,108.1332
Neural HMMs are all you need (for high-quality attention-free TTS)
['Shivam Mehta', 'Éva Székely', 'Jonas Beskow', 'Gustav Eje Henter']
['eess.AS', 'cs.HC', 'cs.LG', 'cs.SD', '68T07', 'I.2.7; I.2.6; G.3; H.5.5']
Neural sequence-to-sequence TTS has achieved significantly better output quality than statistical speech synthesis using HMMs. However, neural TTS is generally not probabilistic and uses non-monotonic attention. Attention failures increase training time and can make synthesis babble incoherently. This paper describes h...
2021-08-30T15:38:00Z
5 pages, 2 figures; final version for ICASSP 2022
null
10.1109/ICASSP43922.2022.9746686
Neural HMMS Are All You Need (For High-Quality Attention-Free TTS)
['Shivam Mehta', 'Éva Székely', 'J. Beskow', 'G. Henter']
2,021
IEEE International Conference on Acoustics, Speech, and Signal Processing
18
50
['Computer Science', 'Engineering']
2,108.13493
Semi-Supervised Exaggeration Detection of Health Science Press Releases
['Dustin Wright', 'Isabelle Augenstein']
['cs.CL', 'cs.LG']
Public trust in science depends on honest and factual communication of scientific papers. However, recent studies have demonstrated a tendency of news media to misrepresent scientific papers by exaggerating their findings. Given this, we present a formalization of and study into the problem of exaggeration detection in...
2021-08-30T19:32:20Z
Accepted to EMNLP 2021; 13 pages, 6 figures, 9 tables
null
null
Semi-Supervised Exaggeration Detection of Health Science Press Releases
['Dustin Wright', 'Isabelle Augenstein']
2,021
Conference on Empirical Methods in Natural Language Processing
13
38
['Computer Science']
2,108.13751
A Search Engine for Discovery of Scientific Challenges and Directions
['Dan Lahav', 'Jon Saad Falcon', 'Bailey Kuehl', 'Sophie Johnson', 'Sravanthi Parasa', 'Noam Shomron', 'Duen Horng Chau', 'Diyi Yang', 'Eric Horvitz', 'Daniel S. Weld', 'Tom Hope']
['cs.CL', 'cs.HC', 'cs.IR']
Keeping track of scientific challenges, advances and emerging directions is a fundamental part of research. However, researchers face a flood of papers that hinders discovery of important knowledge. In biomedicine, this directly impacts human lives. To address this problem, we present a novel task of extraction and sea...
2021-08-31T11:08:20Z
AAAI 2022
AAAI 2022
null
null
null
null
null
null
null
null
2,108.13897
mMARCO: A Multilingual Version of the MS MARCO Passage Ranking Dataset
['Luiz Bonifacio', 'Vitor Jeronymo', 'Hugo Queiroz Abonizio', 'Israel Campiotti', 'Marzieh Fadaee', 'Roberto Lotufo', 'Rodrigo Nogueira']
['cs.CL', 'cs.AI']
The MS MARCO ranking dataset has been widely used for training deep learning models for IR tasks, achieving considerable effectiveness on diverse zero-shot scenarios. However, this type of resource is scarce in languages other than English. In this work, we present mMARCO, a multilingual version of the MS MARCO passage...
2021-08-31T14:53:37Z
null
null
null
mMARCO: A Multilingual Version of the MS MARCO Passage Ranking Dataset
['L. Bonifacio', 'Israel Campiotti', 'R. Lotufo', 'Rodrigo Nogueira']
2,021
null
114
52
['Computer Science']
2,109.00122
FinQA: A Dataset of Numerical Reasoning over Financial Data
['Zhiyu Chen', 'Wenhu Chen', 'Charese Smiley', 'Sameena Shah', 'Iana Borova', 'Dylan Langdon', 'Reema Moussa', 'Matt Beane', 'Ting-Hao Huang', 'Bryan Routledge', 'William Yang Wang']
['cs.CL']
The sheer volume of financial statements makes it difficult for humans to access and analyze a business's financials. Robust numerical reasoning likewise faces unique challenges in this domain. In this work, we focus on answering deep questions over financial data, aiming to automate the analysis of a large corpus of f...
2021-09-01T00:08:14Z
EMNLP 2021
null
null
FinQA: A Dataset of Numerical Reasoning over Financial Data
['Zhiyu Chen', 'Wenhu Chen', 'Charese Smiley', 'Sameena Shah', 'Iana Borova', 'Dylan Langdon', 'Reema Moussa', 'Matthew I. Beane', "Ting-Hao 'Kenneth' Huang", 'Bryan R. Routledge', 'W. Wang']
2,021
Conference on Empirical Methods in Natural Language Processing
356
44
['Computer Science']
2,109.00859
CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for Code Understanding and Generation
['Yue Wang', 'Weishi Wang', 'Shafiq Joty', 'Steven C. H. Hoi']
['cs.CL', 'cs.PL']
Pre-trained models for Natural Languages (NL) like BERT and GPT have been recently shown to transfer well to Programming Languages (PL) and largely benefit a broad set of code-related tasks. Despite their success, most current methods either rely on an encoder-only (or decoder-only) pre-training that is suboptimal for ...
2021-09-02T12:21:06Z
Accepted to EMNLP 2021. 13 pages
null
null
null
null
null
null
null
null
null
2,109.00904
MultiEURLEX -- A multi-lingual and multi-label legal document classification dataset for zero-shot cross-lingual transfer
['Ilias Chalkidis', 'Manos Fergadiotis', 'Ion Androutsopoulos']
['cs.CL']
We introduce MULTI-EURLEX, a new multilingual dataset for topic classification of legal documents. The dataset comprises 65k European Union (EU) laws, officially translated in 23 languages, annotated with multiple labels from the EUROVOC taxonomy. We highlight the effect of temporal concept drift and the importance of ...
2021-09-02T12:52:55Z
9 pages, long paper at EMNLP 2021 proceedings
null
null
null
null
null
null
null
null
null
2,109.01078
Skim-Attention: Learning to Focus via Document Layout
['Laura Nguyen', 'Thomas Scialom', 'Jacopo Staiano', 'Benjamin Piwowarski']
['cs.CL']
Transformer-based pre-training techniques of text and layout have proven effective in a number of document understanding tasks. Despite this success, multimodal pre-training models suffer from very high computational and memory costs. Motivated by human reading strategies, this paper presents Skim-Attention, a new atte...
2021-09-02T16:44:22Z
15 pages, 6 figures, to be published in EMNLP 2021 Findings
null
null
null
null
null
null
null
null
null
2,109.01134
Learning to Prompt for Vision-Language Models
['Kaiyang Zhou', 'Jingkang Yang', 'Chen Change Loy', 'Ziwei Liu']
['cs.CV', 'cs.AI', 'cs.LG']
Large pre-trained vision-language models like CLIP have shown great potential in learning representations that are transferable across a wide range of downstream tasks. Different from the traditional representation learning that is based mostly on discretized labels, vision-language pre-training aligns images and texts...
2021-09-02T17:57:31Z
International Journal of Computer Vision (IJCV), 2022. Update: Adds results on the DOSCO (DOmain Shift in COntext) benchmark
null
10.1007/s11263-022-01653-1
null
null
null
null
null
null
null
2,109.01163
Efficient conformer: Progressive downsampling and grouped attention for automatic speech recognition
['Maxime Burchi', 'Valentin Vielzeuf']
['eess.AS', 'cs.AI', 'cs.CL', 'cs.SD']
The recently proposed Conformer architecture has shown state-of-the-art performances in Automatic Speech Recognition by combining convolution with attention to model both local and global dependencies. In this paper, we study how to reduce the Conformer architecture complexity with a limited computing budget, leading t...
2021-08-31T07:48:06Z
null
ASRU 2021, Dec 2021, Cartagena, Colombia
null
null
null
null
null
null
null
null
2,109.01652
Finetuned Language Models Are Zero-Shot Learners
['Jason Wei', 'Maarten Bosma', 'Vincent Y. Zhao', 'Kelvin Guu', 'Adams Wei Yu', 'Brian Lester', 'Nan Du', 'Andrew M. Dai', 'Quoc V. Le']
['cs.CL']
This paper explores a simple method for improving the zero-shot learning abilities of language models. We show that instruction tuning -- finetuning language models on a collection of tasks described via instructions -- substantially improves zero-shot performance on unseen tasks. We take a 137B parameter pretrained ...
2021-09-03T17:55:52Z
Version 5. Find list of changes in Appendix F (page 35)
null
null
Finetuned Language Models Are Zero-Shot Learners
['Jason Wei', 'Maarten Bosma', 'Vincent Zhao', 'Kelvin Guu', 'Adams Wei Yu', 'Brian Lester', 'Nan Du', 'Andrew M. Dai', 'Quoc V. Le']
2,021
International Conference on Learning Representations
3,814
169
['Computer Science']
2,109.01653
CREAK: A Dataset for Commonsense Reasoning over Entity Knowledge
['Yasumasa Onoe', 'Michael J. Q. Zhang', 'Eunsol Choi', 'Greg Durrett']
['cs.CL', 'cs.AI']
Most benchmark datasets targeting commonsense reasoning focus on everyday scenarios: physical knowledge like knowing that you could fill a cup under a waterfall [Talmor et al., 2019], social knowledge like bumping into someone is awkward [Sap et al., 2019], and other generic situations. However, there is a rich space o...
2021-09-03T17:56:40Z
null
null
null
CREAK: A Dataset for Commonsense Reasoning over Entity Knowledge
['Yasumasa Onoe', 'Michael J.Q. Zhang', 'Eunsol Choi', 'Greg Durrett']
2,021
NeurIPS Datasets and Benchmarks
87
44
['Computer Science']
2,109.01903
Robust fine-tuning of zero-shot models
['Mitchell Wortsman', 'Gabriel Ilharco', 'Jong Wook Kim', 'Mike Li', 'Simon Kornblith', 'Rebecca Roelofs', 'Raphael Gontijo-Lopes', 'Hannaneh Hajishirzi', 'Ali Farhadi', 'Hongseok Namkoong', 'Ludwig Schmidt']
['cs.CV', 'cs.LG']
Large pre-trained models such as CLIP or ALIGN offer consistent accuracy across a range of data distributions when performing zero-shot inference (i.e., without fine-tuning on a specific dataset). Although existing fine-tuning methods substantially improve accuracy on a given target distribution, they often reduce robu...
2021-09-04T17:11:28Z
CVPR 2022
null
null
null
null
null
null
null
null
null
2,109.02492
DialogLM: Pre-trained Model for Long Dialogue Understanding and Summarization
['Ming Zhong', 'Yang Liu', 'Yichong Xu', 'Chenguang Zhu', 'Michael Zeng']
['cs.CL']
Dialogue is an essential part of human communication and cooperation. Existing research mainly focuses on short dialogue scenarios in a one-on-one fashion. However, multi-person interactions in the real world, such as meetings or interviews, are frequently over a few thousand words. There is still a lack of correspondi...
2021-09-06T13:55:03Z
Accepted by AAAI 2022
null
null
null
null
null
null
null
null
null
2,109.02844
Quasinormal modes in two-photon autocorrelation and the geometric-optics approximation
['Wei-Liang Qian', 'Kai Lin', 'Xiao-Mei Kuang', 'Bin Wang', 'Rui-Hong Yue']
['gr-qc']
In this work, we study the black hole light echoes in terms of the two-photon autocorrelation and explore their connection with the quasinormal modes. It is shown that the above time-domain phenomenon can be analyzed by utilizing the well-known frequency-domain relations between the quasinormal modes and characteristic...
2021-09-07T03:55:51Z
12 pages, 3 figures
null
10.1140/epjc/s10052-022-10155-w
null
null
null
null
null
null
null
2,109.02903
IndicBART: A Pre-trained Model for Indic Natural Language Generation
['Raj Dabre', 'Himani Shrotriya', 'Anoop Kunchukuttan', 'Ratish Puduppully', 'Mitesh M. Khapra', 'Pratyush Kumar']
['cs.CL', 'cs.AI']
In this paper, we study pre-trained sequence-to-sequence models for a group of related languages, with a focus on Indic languages. We present IndicBART, a multilingual, sequence-to-sequence pre-trained model focusing on 11 Indic languages and English. IndicBART utilizes the orthographic similarity between Indic scripts...
2021-09-07T07:08:33Z
Published at ACL 2022, 15 pages
null
10.18653/v1/2022.findings-acl.145
IndicBART: A Pre-trained Model for Indic Natural Language Generation
['Raj Dabre', 'Himani Shrotriya', 'Anoop Kunchukuttan', 'Ratish Puduppully', 'Mitesh M. Khapra', 'Pratyush Kumar']
2,021
Findings
74
55
['Computer Science']
2,109.03564
NSP-BERT: A Prompt-based Few-Shot Learner Through an Original Pre-training Task--Next Sentence Prediction
['Yi Sun', 'Yu Zheng', 'Chao Hao', 'Hangping Qiu']
['cs.CL', 'cs.AI']
Using prompts to utilize language models to perform various downstream tasks, also known as prompt-based learning or prompt-learning, has lately gained significant success in comparison to the pre-train and fine-tune paradigm. Nonetheless, virtually all prompt-based methods are token-level, meaning they all utilize GPT...
2021-09-08T11:57:08Z
Published at COLING2022, long paper
null
null
null
null
null
null
null
null
null
2,109.0357
Biomedical and Clinical Language Models for Spanish: On the Benefits of Domain-Specific Pretraining in a Mid-Resource Scenario
['Casimiro Pio Carrino', 'Jordi Armengol-Estapé', 'Asier Gutiérrez-Fandiño', 'Joan Llop-Palao', 'Marc Pàmies', 'Aitor Gonzalez-Agirre', 'Marta Villegas']
['cs.CL']
This work presents biomedical and clinical language models for Spanish by experimenting with different pretraining choices, such as masking at word and subword level, varying the vocabulary size and testing with domain data, looking for better language representations. Interestingly, in the absence of enough clinical d...
2021-09-08T12:12:07Z
9 pages
null
null
Biomedical and Clinical Language Models for Spanish: On the Benefits of Domain-Specific Pretraining in a Mid-Resource Scenario
['C. Carrino', "Jordi Armengol-Estap'e", 'Asier Gutiérrez-Fandiño', 'Joan Llop-Palao', 'Marc Pàmies', 'Aitor Gonzalez-Agirre', 'Marta Villegas']
2,021
arXiv.org
44
34
['Computer Science']
2,109.03814
Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with Transformers
['Zhiqi Li', 'Wenhai Wang', 'Enze Xie', 'Zhiding Yu', 'Anima Anandkumar', 'Jose M. Alvarez', 'Ping Luo', 'Tong Lu']
['cs.CV']
Panoptic segmentation involves a combination of joint semantic segmentation and instance segmentation, where image contents are divided into two types: things and stuff. We present Panoptic SegFormer, a general framework for panoptic segmentation with transformers. It contains three innovative components: an efficient ...
2021-09-08T17:59:12Z
Accepted to CVPR 2022
null
null
null
null
null
null
null
null
null
2,109.03955
NU:BRIEF -- A Privacy-aware Newsletter Personalization Engine for Publishers
['Ernesto Diaz-Aviles', 'Claudia Orellana-Rodriguez', 'Igor Brigadir', 'Reshma Narayanan Kutty']
['cs.DL', 'cs.CY', 'cs.HC', 'cs.IR', 'cs.LG']
Newsletters have (re-) emerged as a powerful tool for publishers to engage with their readers directly and more effectively. Despite the diversity in their audiences, publishers' newsletters remain largely a one-size-fits-all offering, which is suboptimal. In this paper, we present NU:BRIEF, a web application for publi...
2021-09-08T22:36:05Z
Fifteenth ACM Conference on Recommender Systems (RecSys '21), September 27-October 1, 2021, Amsterdam, Netherlands
null
10.1145/3460231.3478884
null
null
null
null
null
null
null
2,109.04127
Word-Level Coreference Resolution
['Vladimir Dobrovolskii']
['cs.CL', 'I.2.7']
Recent coreference resolution models rely heavily on span representations to find coreference links between word spans. As the number of spans is $O(n^2)$ in the length of text and the number of potential links is $O(n^4)$, various pruning techniques are necessary to make this approach computationally feasible. We prop...
2021-09-09T09:26:02Z
Accepted to EMNLP-2021
In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing (pp. 7670-7675). Association for Computational Linguistics 2021
10.18653/v1/2021.emnlp-main.605
Word-Level Coreference Resolution
['V. Dobrovolskii']
2,021
Conference on Empirical Methods in Natural Language Processing
74
22
['Computer Science']
2,109.04263
Complex chiral columns made of achiral quinoxaline derivatives with semi-flexible cores
['Paulina Rybak', 'Adam Krowczynski', 'Jadwiga Szydlowska', 'Damian Pociecha', 'Ewa Gorecka']
['cond-mat.mtrl-sci', 'cond-mat.soft']
Mesogenic materials, quinoxaline derivatives with semi-flexible cores, are reported to form new type of 3D columnar structure with large crystallographic unit cell and Fddd symmetry below columnar hexagonal phase. The 3D columnar structure is a result of frustration imposed by arrangement of helical columns of opposite...
2021-09-09T13:33:22Z
null
null
null
null
null
null
null
null
null
null
2,109.04607
IndoBERTweet: A Pretrained Language Model for Indonesian Twitter with Effective Domain-Specific Vocabulary Initialization
['Fajri Koto', 'Jey Han Lau', 'Timothy Baldwin']
['cs.CL']
We present IndoBERTweet, the first large-scale pretrained model for Indonesian Twitter that is trained by extending a monolingually-trained Indonesian BERT model with additive domain-specific vocabulary. We focus in particular on efficient model adaptation under vocabulary mismatch, and benchmark different ways of init...
2021-09-10T01:27:51Z
Accepted at EMNLP 2021
null
null
IndoBERTweet: A Pretrained Language Model for Indonesian Twitter with Effective Domain-Specific Vocabulary Initialization
['Fajri Koto', 'Jey Han Lau', 'Timothy Baldwin']
2,021
Conference on Empirical Methods in Natural Language Processing
85
38
['Computer Science']
2,109.0465
What Changes Can Large-scale Language Models Bring? Intensive Study on HyperCLOVA: Billions-scale Korean Generative Pretrained Transformers
['Boseop Kim', 'HyoungSeok Kim', 'Sang-Woo Lee', 'Gichang Lee', 'Donghyun Kwak', 'Dong Hyeon Jeon', 'Sunghyun Park', 'Sungju Kim', 'Seonhoon Kim', 'Dongpil Seo', 'Heungsub Lee', 'Minyoung Jeong', 'Sungjae Lee', 'Minsub Kim', 'Suk Hyun Ko', 'Seokhun Kim', 'Taeyong Park', 'Jinuk Kim', 'Soyoung Kang', 'Na-Hyeon Ryu', 'Kan...
['cs.CL']
GPT-3 shows remarkable in-context learning ability of large-scale language models (LMs) trained on hundreds of billion scale data. Here we address some remaining issues less reported by the GPT-3 paper, such as a non-English LM, the performances of different sized models, and the effect of recently introduced prompt op...
2021-09-10T03:32:19Z
Accepted to EMNLP2021 as a long paper. Fixed some typos
null
null
What Changes Can Large-scale Language Models Bring? Intensive Study on HyperCLOVA: Billions-scale Korean Generative Pretrained Transformers
['Boseop Kim', 'Hyoungseok Kim', 'Sang-Woo Lee', 'Gichang Lee', 'Donghyun Kwak', 'D. Jeon', 'Sunghyun Park', 'Sungju Kim', 'Seonhoon Kim', 'D. Seo', 'Heungsub Lee', 'Minyoung Jeong', 'Sungjae Lee', 'Minsub Kim', 'SukHyun Ko', 'Seokhun Kim', 'Taeyong Park', 'Jinuk Kim', 'Soyoung Kang', 'Nahyeon Ryu', 'Kang Min Yoo', 'Mi...
2,021
Conference on Empirical Methods in Natural Language Processing
124
56
['Computer Science']
2,109.04655
Zero-Shot Dialogue State Tracking via Cross-Task Transfer
['Zhaojiang Lin', 'Bing Liu', 'Andrea Madotto', 'Seungwhan Moon', 'Paul Crook', 'Zhenpeng Zhou', 'Zhiguang Wang', 'Zhou Yu', 'Eunjoon Cho', 'Rajen Subba', 'Pascale Fung']
['cs.CL']
Zero-shot transfer learning for dialogue state tracking (DST) enables us to handle a variety of task-oriented dialogue domains without the expense of collecting in-domain data. In this work, we propose to transfer the \textit{cross-task} knowledge from general question answering (QA) corpora for the zero-shot DST task....
2021-09-10T03:57:56Z
EMNLP 2021
null
null
null
null
null
null
null
null
null
2,109.04689
Generating Self-Contained and Summary-Centric Question Answer Pairs via Differentiable Reward Imitation Learning
['Li Zhou', 'Kevin Small', 'Yong Zhang', 'Sandeep Atluri']
['cs.CL', 'cs.AI', 'cs.LG']
Motivated by suggested question generation in conversational news recommendation systems, we propose a model for generating question-answer pairs (QA pairs) with self-contained, summary-centric questions and length-constrained, article-summarizing answers. We begin by collecting a new dataset of news articles with ques...
2021-09-10T06:34:55Z
To appear in Proceedings of EMNLP 2021
null
null
null
null
null
null
null
null
null
2,109.04711
Pre-train or Annotate? Domain Adaptation with a Constrained Budget
['Fan Bai', 'Alan Ritter', 'Wei Xu']
['cs.CL']
Recent work has demonstrated that pre-training in-domain language models can boost performance when adapting to a new domain. However, the costs associated with pre-training raise an important question: given a fixed budget, what steps should an NLP practitioner take to maximize performance? In this paper, we view doma...
2021-09-10T07:28:26Z
Accepted to EMNLP 2021
null
null
null
null
null
null
null
null
null
2,109.04838
Block Pruning For Faster Transformers
['François Lagunas', 'Ella Charlaix', 'Victor Sanh', 'Alexander M. Rush']
['cs.LG', 'cs.CL', 'I.2.6; I.2.7']
Pre-training has improved model accuracy for both classification and generation tasks at the cost of introducing much larger and slower models. Pruning methods have proven to be an effective way of reducing model size, whereas distillation methods are proven for speeding up inference. We introduce a block pruning appro...
2021-09-10T12:46:32Z
EMNLP 2021. Code, hyper-parameters, evaluation results and checkpoints available at https://github.com/huggingface/nn_pruning
null
null
null
null
null
null
null
null
null
2,109.05014
An Empirical Study of GPT-3 for Few-Shot Knowledge-Based VQA
['Zhengyuan Yang', 'Zhe Gan', 'Jianfeng Wang', 'Xiaowei Hu', 'Yumao Lu', 'Zicheng Liu', 'Lijuan Wang']
['cs.CV']
Knowledge-based visual question answering (VQA) involves answering questions that require external knowledge not present in the image. Existing methods first retrieve knowledge from external resources, then reason over the selected knowledge, the input image, and question for answer prediction. However, this two-step a...
2021-09-10T17:51:06Z
AAAI 2022 (Oral Presentation)
null
null
null
null
null
null
null
null
null
2,109.0507
Instance-Conditioned GAN
['Arantxa Casanova', 'Marlène Careil', 'Jakob Verbeek', 'Michal Drozdzal', 'Adriana Romero-Soriano']
['cs.CV', 'cs.LG']
Generative Adversarial Networks (GANs) can generate near photo realistic images in narrow domains such as human faces. Yet, modeling complex distributions of datasets such as ImageNet and COCO-Stuff remains challenging in unconditional settings. In this paper, we take inspiration from kernel density estimation techniqu...
2021-09-10T19:08:45Z
Accepted at NeurIPS2021
null
null
Instance-Conditioned GAN
['Arantxa Casanova', 'Marlene Careil', 'Jakob Verbeek', 'M. Drozdzal', 'Adriana Romero-Soriano']
2,021
Neural Information Processing Systems
138
59
['Computer Science']
2,109.05093
PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models
['Torsten Scholak', 'Nathan Schucher', 'Dzmitry Bahdanau']
['cs.CL', 'cs.PL']
Large pre-trained language models for textual data have an unconstrained output space; at each decoding step, they can produce any of 10,000s of sub-word tokens. When fine-tuned to target constrained formal languages like SQL, these models often generate invalid code, rendering it unusable. We propose PICARD (code and ...
2021-09-10T20:14:08Z
Accepted to EMNLP 2021. 7 pages
null
null
PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models
['Torsten Scholak', 'Nathan Schucher', 'Dzmitry Bahdanau']
2,021
Conference on Empirical Methods in Natural Language Processing
397
20
['Computer Science']
2,109.05153
Natural SQL: Making SQL Easier to Infer from Natural Language Specifications
['Yujian Gan', 'Xinyun Chen', 'Jinxia Xie', 'Matthew Purver', 'John R. Woodward', 'John Drake', 'Qiaofu Zhang']
['cs.CL']
Addressing the mismatch between natural language descriptions and the corresponding SQL queries is a key challenge for text-to-SQL translation. To bridge this gap, we propose an SQL intermediate representation (IR) called Natural SQL (NatSQL). Specifically, NatSQL preserves the core functionalities of SQL, while it sim...
2021-09-11T01:53:55Z
To appear in EMNLP Findings 2021
null
null
Natural SQL: Making SQL Easier to Infer from Natural Language Specifications
['Yujian Gan', 'Xinyun Chen', 'Jinxia Xie', 'Matthew Purver', 'J. Woodward', 'J. Drake', 'Qiaofu Zhang']
2,021
Conference on Empirical Methods in Natural Language Processing
95
32
['Computer Science']
2,109.05217
Empirical Analysis of Training Strategies of Transformer-based Japanese Chit-chat Systems
['Hiroaki Sugiyama', 'Masahiro Mizukami', 'Tsunehiro Arimoto', 'Hiromi Narimatsu', 'Yuya Chiba', 'Hideharu Nakajima', 'Toyomi Meguro']
['cs.CL', 'cs.AI']
In recent years, several high-performance conversational systems have been proposed based on the Transformer encoder-decoder model. Although previous studies analyzed the effects of the model parameters and the decoding method on subjective dialogue evaluations with overall metrics, they did not analyze how the differe...
2021-09-11T08:24:23Z
11 pages, 2 figures
null
null
null
null
null
null
null
null
null
2,109.0546
End-to-End Conversational Search for Online Shopping with Utterance Transfer
['Liqiang Xiao', 'Jun Ma2', 'Xin Luna Dong', 'Pascual Martinez-Gomez', 'Nasser Zalmout', 'Wei Chen', 'Tong Zhao', 'Hao He', 'Yaohui Jin']
['cs.CL', 'cs.AI']
Successful conversational search systems can present natural, adaptive and interactive shopping experience for online shopping customers. However, building such systems from scratch faces real word challenges from both imperfect product schema/knowledge and lack of training dialog data.In this work we first propose Con...
2021-09-12T08:33:44Z
null
null
null
End-to-End Conversational Search for Online Shopping with Utterance Transfer
['Liqiang Xiao', 'Jun Ma', 'Xin Dong', 'Pascual Martínez-Gómez', 'Nasser Zalmout', 'Wei Chen', 'Tong Zhao', 'Hao He', 'Yaohui Jin']
2,021
Conference on Empirical Methods in Natural Language Processing
12
29
['Computer Science']
2,109.05729
CPT: A Pre-Trained Unbalanced Transformer for Both Chinese Language Understanding and Generation
['Yunfan Shao', 'Zhichao Geng', 'Yitao Liu', 'Junqi Dai', 'Hang Yan', 'Fei Yang', 'Li Zhe', 'Hujun Bao', 'Xipeng Qiu']
['cs.CL']
In this paper, we take the advantage of previous pre-trained models (PTMs) and propose a novel Chinese Pre-trained Unbalanced Transformer (CPT). Different from previous Chinese PTMs, CPT is designed to utilize the shared knowledge between natural language understanding (NLU) and natural language generation (NLG) to boo...
2021-09-13T06:25:45Z
Code is available at https://github.com/fastnlp/CPT
null
10.1007/s11432-021-3536-5
CPT: A Pre-Trained Unbalanced Transformer for Both Chinese Language Understanding and Generation
['Yunfan Shao', 'Zhichao Geng', 'Yitao Liu', 'Junqi Dai', 'Fei Yang', 'Li Zhe', 'H. Bao', 'Xipeng Qiu']
2,021
Science China Information Sciences
152
46
['Computer Science']
2,109.06304
Phrase-BERT: Improved Phrase Embeddings from BERT with an Application to Corpus Exploration
['Shufan Wang', 'Laure Thompson', 'Mohit Iyyer']
['cs.CL']
Phrase representations derived from BERT often do not exhibit complex phrasal compositionality, as the model relies instead on lexical similarity to determine semantic relatedness. In this paper, we propose a contrastive fine-tuning objective that enables BERT to produce more powerful phrase embeddings. Our approach (P...
2021-09-13T20:31:57Z
EMNLP 2021 Conference Camera Ready
null
null
Phrase-BERT: Improved Phrase Embeddings from BERT with an Application to Corpus Exploration
['Shufan Wang', 'Laure Thompson', 'Mohit Iyyer']
2,021
Conference on Empirical Methods in Natural Language Processing
68
49
['Computer Science']
2,109.06379
Compression, Transduction, and Creation: A Unified Framework for Evaluating Natural Language Generation
['Mingkai Deng', 'Bowen Tan', 'Zhengzhong Liu', 'Eric P. Xing', 'Zhiting Hu']
['cs.CL', 'cs.LG']
Natural language generation (NLG) spans a broad range of tasks, each of which serves for specific objectives and desires different properties of generated text. The complexity makes automatic evaluation of NLG particularly challenging. Previous work has typically focused on a single task and developed individual evalua...
2021-09-14T01:00:42Z
EMNLP 2021, Code available at https://github.com/tanyuqian/ctc-gen-eval
null
null
null
null
null
null
null
null
null
2,109.06402
Exploring Personality and Online Social Engagement: An Investigation of MBTI Users on Twitter
['Partha Kadambi']
['cs.CL', 'cs.LG']
Text-based personality prediction by computational models is an emerging field with the potential to significantly improve on key weaknesses of survey-based personality assessment. We investigate 3848 profiles from Twitter with self-labeled Myers-Briggs personality traits (MBTI) - a framework closely related to the Fiv...
2021-09-14T02:26:30Z
null
null
null
null
null
null
null
null
null
null
2,109.0687
Performance-Efficiency Trade-offs in Unsupervised Pre-training for Speech Recognition
['Felix Wu', 'Kwangyoun Kim', 'Jing Pan', 'Kyu Han', 'Kilian Q. Weinberger', 'Yoav Artzi']
['cs.CL', 'cs.LG', 'cs.SD', 'eess.AS']
This paper is a study of performance-efficiency trade-offs in pre-trained models for automatic speech recognition (ASR). We focus on wav2vec 2.0, and formalize several architecture designs that influence both the model performance and its efficiency. Putting together all our observations, we introduce SEW (Squeezed and...
2021-09-14T17:58:09Z
Code available at https://github.com/asappresearch/sew
null
null
Performance-Efficiency Trade-Offs in Unsupervised Pre-Training for Speech Recognition
['Felix Wu', 'Kwangyoun Kim', 'Jing Pan', 'Kyu J. Han', 'Kilian Q. Weinberger', 'Yoav Artzi']
2,021
IEEE International Conference on Acoustics, Speech, and Signal Processing
75
73
['Computer Science', 'Engineering']
2,109.06912
fairseq S^2: A Scalable and Integrable Speech Synthesis Toolkit
['Changhan Wang', 'Wei-Ning Hsu', 'Yossi Adi', 'Adam Polyak', 'Ann Lee', 'Peng-Jen Chen', 'Jiatao Gu', 'Juan Pino']
['eess.AS', 'cs.CL', 'cs.SD']
This paper presents fairseq S^2, a fairseq extension for speech synthesis. We implement a number of autoregressive (AR) and non-AR text-to-speech models, and their multi-speaker variants. To enable training speech synthesis models with less curated data, a number of preprocessing tools are built and their importance is...
2021-09-14T18:20:28Z
Accepted to EMNLP 2021 Demo
null
null
null
null
null
null
null
null
null
2,109.07161
Resolution-robust Large Mask Inpainting with Fourier Convolutions
['Roman Suvorov', 'Elizaveta Logacheva', 'Anton Mashikhin', 'Anastasia Remizova', 'Arsenii Ashukha', 'Aleksei Silvestrov', 'Naejin Kong', 'Harshith Goka', 'Kiwoong Park', 'Victor Lempitsky']
['cs.CV', 'eess.IV']
Modern image inpainting systems, despite the significant progress, often struggle with large missing areas, complex geometric structures, and high-resolution images. We find that one of the main reasons for that is the lack of an effective receptive field in both the inpainting network and the loss function. To allevia...
2021-09-15T08:54:29Z
Winter Conference on Applications of Computer Vision (WACV 2022)
null
null
null
null
null
null
null
null
null
2,109.07306
Allocating Large Vocabulary Capacity for Cross-lingual Language Model Pre-training
['Bo Zheng', 'Li Dong', 'Shaohan Huang', 'Saksham Singhal', 'Wanxiang Che', 'Ting Liu', 'Xia Song', 'Furu Wei']
['cs.CL']
Compared to monolingual models, cross-lingual models usually require a more expressive vocabulary to represent all languages adequately. We find that many languages are under-represented in recent cross-lingual language models due to the limited vocabulary capacity. To this end, we propose an algorithm VoCap to determi...
2021-09-15T14:04:16Z
EMNLP 2021
null
null
Allocating Large Vocabulary Capacity for Cross-Lingual Language Model Pre-Training
['Bo Zheng', 'Li Dong', 'Shaohan Huang', 'Saksham Singhal', 'Wanxiang Che', 'Ting Liu', 'Xia Song', 'Furu Wei']
2,021
Conference on Empirical Methods in Natural Language Processing
22
43
['Computer Science']
2,109.07765
Spanish Biomedical Crawled Corpus: A Large, Diverse Dataset for Spanish Biomedical Language Models
['Casimiro Pio Carrino', 'Jordi Armengol-Estapé', 'Ona de Gibert Bonet', 'Asier Gutiérrez-Fandiño', 'Aitor Gonzalez-Agirre', 'Martin Krallinger', 'Marta Villegas']
['cs.CL']
We introduce CoWeSe (the Corpus Web Salud Espa\~nol), the largest Spanish biomedical corpus to date, consisting of 4.5GB (about 750M tokens) of clean plain text. CoWeSe is the result of a massive crawler on 3000 Spanish domains executed in 2020. The corpus is openly available and already preprocessed. CoWeSe is an impo...
2021-09-16T07:22:28Z
null
null
null
null
null
null
null
null
null
null
2,109.07958
TruthfulQA: Measuring How Models Mimic Human Falsehoods
['Stephanie Lin', 'Jacob Hilton', 'Owain Evans']
['cs.CL', 'cs.AI', 'cs.CY', 'cs.LG']
We propose a benchmark to measure whether a language model is truthful in generating answers to questions. The benchmark comprises 817 questions that span 38 categories, including health, law, finance and politics. We crafted questions that some humans would answer falsely due to a false belief or misconception. To per...
2021-09-08T17:15:27Z
ACL 2022 (main conference); the TruthfulQA benchmark and evaluation code is available at https://github.com/sylinrl/TruthfulQA
null
null
null
null
null
null
null
null
null
2,109.08079
Context-NER : Contextual Phrase Generation at Scale
['Himanshu Gupta', 'Shreyas Verma', 'Santosh Mashetty', 'Swaroop Mishra']
['cs.IR', 'cs.CL', 'cs.LG']
Named Entity Recognition (NER) has seen significant progress in recent years, with numerous state-of-the-art (SOTA) models achieving high performance. However, very few studies have focused on the generation of entities' context. In this paper, we introduce CONTEXT-NER, a task that aims to generate the relevant context...
2021-09-16T16:10:05Z
29 pages, 5 Figures, 2 AlgorithmS, 17 Tables. Accepted in NeurIPS 2022 - Efficient Natural Language and Speech Processing (ENLSP) Workshop
null
null
null
null
null
null
null
null
null
2,109.08203
Torch.manual_seed(3407) is all you need: On the influence of random seeds in deep learning architectures for computer vision
['David Picard']
['cs.CV']
In this paper I investigate the effect of random seed selection on the accuracy when using popular deep learning architectures for computer vision. I scan a large amount of seeds (up to $10^4$) on CIFAR 10 and I also scan fewer seeds on Imagenet using pre-trained models to investigate large scale datasets. The conclusi...
2021-09-16T20:10:12Z
fixed typos
null
null
null
null
null
null
null
null
null
2,109.08238
Habitat-Matterport 3D Dataset (HM3D): 1000 Large-scale 3D Environments for Embodied AI
['Santhosh K. Ramakrishnan', 'Aaron Gokaslan', 'Erik Wijmans', 'Oleksandr Maksymets', 'Alex Clegg', 'John Turner', 'Eric Undersander', 'Wojciech Galuba', 'Andrew Westbury', 'Angel X. Chang', 'Manolis Savva', 'Yili Zhao', 'Dhruv Batra']
['cs.CV', 'cs.AI']
We present the Habitat-Matterport 3D (HM3D) dataset. HM3D is a large-scale dataset of 1,000 building-scale 3D reconstructions from a diverse set of real-world locations. Each scene in the dataset consists of a textured 3D mesh reconstruction of interiors such as multi-floor residences, stores, and other private indoor ...
2021-09-16T22:01:24Z
21 pages, 14 figures
null
null
null
null
null
null
null
null
null
2,109.08564
Slot Filling for Biomedical Information Extraction
['Yannis Papanikolaou', 'Marlene Staib', 'Justin Grace', 'Francine Bennett']
['cs.CL', 'cs.IR', 'cs.LG']
Information Extraction (IE) from text refers to the task of extracting structured knowledge from unstructured text. The task typically consists of a series of sub-tasks such as Named Entity Recognition and Relation Extraction. Sourcing entity and relation type specific training data is a major bottleneck in domains wit...
2021-09-17T14:16:00Z
null
null
null
null
null
null
null
null
null
null
2,109.08914
Text Detoxification using Large Pre-trained Neural Models
['David Dale', 'Anton Voronov', 'Daryna Dementieva', 'Varvara Logacheva', 'Olga Kozlova', 'Nikita Semenov', 'Alexander Panchenko']
['cs.CL', 'cs.LG']
We present two novel unsupervised methods for eliminating toxicity in text. Our first method combines two recent ideas: (1) guidance of the generation process with small style-conditional language models and (2) use of paraphrasing models to perform style transfer. We use a well-performing paraphraser guided by style-t...
2021-09-18T11:55:32Z
Accepted to the EMNLP 2021 conference
null
null
null
null
null
null
null
null
null
2,109.09209
CLIFF: Contrastive Learning for Improving Faithfulness and Factuality in Abstractive Summarization
['Shuyang Cao', 'Lu Wang']
['cs.CL']
We study generating abstractive summaries that are faithful and factually consistent with the given articles. A novel contrastive learning formulation is presented, which leverages both reference summaries, as positive training data, and automatically generated erroneous summaries, as negative training data, to train s...
2021-09-19T20:05:21Z
EMNLP 2021
null
null
CLIFF: Contrastive Learning for Improving Faithfulness and Factuality in Abstractive Summarization
['Shuyang Cao', 'Lu Wang']
2,021
Conference on Empirical Methods in Natural Language Processing
182
58
['Computer Science']
2,109.09667
On Generalization in Coreference Resolution
['Shubham Toshniwal', 'Patrick Xia', 'Sam Wiseman', 'Karen Livescu', 'Kevin Gimpel']
['cs.CL']
While coreference resolution is defined independently of dataset domain, most models for performing coreference resolution do not transfer well to unseen domains. We consolidate a set of 8 coreference resolution datasets targeting different domains to evaluate the off-the-shelf performance of models. We then mix three ...
2021-09-20T16:33:22Z
CRAC 2021
null
null
null
null
null
null
null
null
null
2,109.09701
BARTpho: Pre-trained Sequence-to-Sequence Models for Vietnamese
['Nguyen Luong Tran', 'Duong Minh Le', 'Dat Quoc Nguyen']
['cs.CL']
We present BARTpho with two versions, BARTpho-syllable and BARTpho-word, which are the first public large-scale monolingual sequence-to-sequence models pre-trained for Vietnamese. BARTpho uses the "large" architecture and the pre-training scheme of the sequence-to-sequence denoising autoencoder BART, thus it is especia...
2021-09-20T17:14:22Z
In Proceedings of INTERSPEECH 2022 (to appear)
null
null
null
null
null
null
null
null
null
2,109.10086
SPLADE v2: Sparse Lexical and Expansion Model for Information Retrieval
['Thibault Formal', 'Carlos Lassance', 'Benjamin Piwowarski', 'Stéphane Clinchant']
['cs.IR', 'cs.AI', 'cs.CL']
In neural Information Retrieval (IR), ongoing research is directed towards improving the first retriever in ranking pipelines. Learning dense embeddings to conduct retrieval using efficient approximate nearest neighbors methods has proven to work well. Meanwhile, there has been a growing interest in learning \emph{spar...
2021-09-21T10:43:42Z
5 pages. arXiv admin note: substantial text overlap with arXiv:2107.05720
null
null
null
null
null
null
null
null
null
2,109.10282
TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models
['Minghao Li', 'Tengchao Lv', 'Jingye Chen', 'Lei Cui', 'Yijuan Lu', 'Dinei Florencio', 'Cha Zhang', 'Zhoujun Li', 'Furu Wei']
['cs.CL', 'cs.CV']
Text recognition is a long-standing research problem for document digitalization. Existing approaches are usually built based on CNN for image understanding and RNN for char-level text generation. In addition, another language model is usually needed to improve the overall accuracy as a post-processing step. In this pa...
2021-09-21T16:01:56Z
Work in Progress
null
null
null
null
null
null
null
null
null
2,109.10686
Scale Efficiently: Insights from Pre-training and Fine-tuning Transformers
['Yi Tay', 'Mostafa Dehghani', 'Jinfeng Rao', 'William Fedus', 'Samira Abnar', 'Hyung Won Chung', 'Sharan Narang', 'Dani Yogatama', 'Ashish Vaswani', 'Donald Metzler']
['cs.CL', 'cs.AI', 'cs.CV', 'cs.LG']
There remain many open questions pertaining to the scaling behaviour of Transformer architectures. These scaling decisions and findings can be critical, as training runs often come with an associated computational cost which have both financial and/or environmental impact. The goal of this paper is to present scaling i...
2021-09-22T12:29:15Z
ICLR 2022 + Updated Checkpoint Release
null
null
null
null
null
null
null
null
null
2,109.11314
ParaShoot: A Hebrew Question Answering Dataset
['Omri Keren', 'Omer Levy']
['cs.CL']
NLP research in Hebrew has largely focused on morphology and syntax, where rich annotated datasets in the spirit of Universal Dependencies are available. Semantic datasets, however, are in short supply, hindering crucial advances in the development of NLP technology in Hebrew. In this work, we present ParaShoot, the fi...
2021-09-23T11:59:38Z
null
null
null
ParaShoot: A Hebrew Question Answering Dataset
['Omri Keren', 'Omer Levy']
2,021
Workshop on Machine Reading for Question Answering
17
20
['Computer Science']
2,109.1168
Simple and Effective Zero-shot Cross-lingual Phoneme Recognition
['Qiantong Xu', 'Alexei Baevski', 'Michael Auli']
['cs.CL', 'cs.LG', 'cs.SD']
Recent progress in self-training, self-supervised pretraining and unsupervised learning enabled well performing speech recognition systems without any labeled data. However, in many cases there is labeled data available for related languages which is not utilized by these methods. This paper extends previous work on ze...
2021-09-23T22:50:32Z
null
null
null
null
null
null
null
null
null
null
2,109.12068
AraT5: Text-to-Text Transformers for Arabic Language Generation
['El Moatez Billah Nagoudi', 'AbdelRahim Elmadany', 'Muhammad Abdul-Mageed']
['cs.CL']
Transfer learning with a unified Transformer framework (T5) that converts all language problems into a text-to-text format was recently proposed as a simple and effective transfer learning approach. Although a multilingual version of the T5 model (mT5) was also introduced, it is not clear how well it can fare on non-En...
2021-08-31T02:02:10Z
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (ACL 2022). All authors contributed equally
null
null
null
null
null
null
null
null
null
2,109.12346
DziriBERT: a Pre-trained Language Model for the Algerian Dialect
['Amine Abdaoui', 'Mohamed Berrimi', 'Mourad Oussalah', 'Abdelouahab Moussaoui']
['cs.CL', 'cs.LG']
Pre-trained transformers are now the de facto models in Natural Language Processing given their state-of-the-art results in many tasks and languages. However, most of the current models have been trained on languages for which large text resources are already available (such as English, French, Arabic, etc.). Therefore...
2021-09-25T11:51:35Z
4 Pages
null
null
DziriBERT: a Pre-trained Language Model for the Algerian Dialect
['Amine Abdaoui', 'Mohamed Berrimi', 'M. Oussalah', 'A. Moussaoui']
2,021
arXiv.org
45
27
['Computer Science']
2,109.12848
A General Gaussian Heatmap Label Assignment for Arbitrary-Oriented Object Detection
['Zhanchao Huang', 'Wei Li', 'Xiang-Gen Xia', 'Ran Tao']
['cs.CV']
Recently, many arbitrary-oriented object detection (AOOD) methods have been proposed and attracted widespread attention in many fields. However, most of them are based on anchor-boxes or standard Gaussian heatmaps. Such label assignment strategy may not only fail to reflect the shape and direction characteristics of ar...
2021-09-27T07:46:09Z
16 pages, 13 figures
IEEE Transactions on Image Processing 2022
10.1109/TIP.2022.3148874
A General Gaussian Heatmap Label Assignment for Arbitrary-Oriented Object Detection
['Zhanchao Huang', 'Wei Li', 'X. Xia', 'R. Tao']
2,021
IEEE Transactions on Image Processing
99
48
['Medicine', 'Computer Science']
2,109.1287
MFAQ: a Multilingual FAQ Dataset
['Maxime De Bruyn', 'Ehsan Lotfi', 'Jeska Buhmann', 'Walter Daelemans']
['cs.CL']
In this paper, we present the first multilingual FAQ dataset publicly available. We collected around 6M FAQ pairs from the web, in 21 different languages. Although this is significantly larger than existing FAQ retrieval datasets, it comes with its own challenges: duplication of content and uneven distribution of topic...
2021-09-27T08:43:25Z
Accepted at MRQA workshop (EMNLP 2021)
null
null
null
null
null
null
null
null
null
2,109.13059
Trans-Encoder: Unsupervised sentence-pair modelling through self- and mutual-distillations
['Fangyu Liu', 'Yunlong Jiao', 'Jordan Massiah', 'Emine Yilmaz', 'Serhii Havrylov']
['cs.CL', 'cs.AI', 'cs.LG']
In NLP, a large volume of tasks involve pairwise comparison between two sequences (e.g. sentence similarity and paraphrase identification). Predominantly, two formulations are used for sentence-pair tasks: bi-encoders and cross-encoders. Bi-encoders produce fixed-dimensional sentence representations and are computation...
2021-09-27T14:06:47Z
ICLR 2022; code and models are released at https://github.com/amzn/trans-encoder
null
null
Trans-Encoder: Unsupervised sentence-pair modelling through self- and mutual-distillations
['Fangyu Liu', 'Serhii Havrylov', 'Yunlong Jiao', 'Jordan Massiah', 'Emine Yilmaz']
2,021
International Conference on Learning Representations
29
39
['Computer Science']
2,109.13228
PASS: An ImageNet replacement for self-supervised pretraining without humans
['Yuki M. Asano', 'Christian Rupprecht', 'Andrew Zisserman', 'Andrea Vedaldi']
['cs.CV', 'cs.CY']
Computer vision has long relied on ImageNet and other large datasets of images sampled from the Internet for pretraining models. However, these datasets have ethical and technical shortcomings, such as containing personal information taken without consent, unclear license usage, biases, and, in some cases, even problem...
2021-09-27T17:59:39Z
Accepted to NeurIPS Track on Datasets and Benchmarks 2021. Webpage: https://www.robots.ox.ac.uk/~vgg/research/pass/
null
null
PASS: An ImageNet replacement for self-supervised pretraining without humans
['Yuki M. Asano', 'C. Rupprecht', 'Andrew Zisserman', 'A. Vedaldi']
2,021
NeurIPS Datasets and Benchmarks
58
112
['Computer Science']
2,109.13821
Diffusion-Based Voice Conversion with Fast Maximum Likelihood Sampling Scheme
['Vadim Popov', 'Ivan Vovk', 'Vladimir Gogoryan', 'Tasnima Sadekova', 'Mikhail Kudinov', 'Jiansheng Wei']
['cs.SD', 'cs.LG', 'stat.ML']
Voice conversion is a common speech synthesis task which can be solved in different ways depending on a particular real-world scenario. The most challenging one often referred to as one-shot many-to-many voice conversion consists in copying the target voice from only one reference utterance in the most general case whe...
2021-09-28T15:48:22Z
null
null
null
Diffusion-Based Voice Conversion with Fast Maximum Likelihood Sampling Scheme
['Vadim Popov', 'Ivan Vovk', 'Vladimir Gogoryan', 'Tasnima Sadekova', 'Mikhail Kudinov', 'Jiansheng Wei']
2,021
International Conference on Learning Representations
136
46
['Computer Science', 'Mathematics']
2,109.15099
PP-LCNet: A Lightweight CPU Convolutional Neural Network
['Cheng Cui', 'Tingquan Gao', 'Shengyu Wei', 'Yuning Du', 'Ruoyu Guo', 'Shuilong Dong', 'Bin Lu', 'Ying Zhou', 'Xueying Lv', 'Qiwen Liu', 'Xiaoguang Hu', 'Dianhai Yu', 'Yanjun Ma']
['cs.CV']
We propose a lightweight CPU network based on the MKLDNN acceleration strategy, named PP-LCNet, which improves the performance of lightweight models on multiple tasks. This paper lists technologies which can improve network accuracy while the latency is almost constant. With these improvements, the accuracy of PP-LCNet...
2021-09-17T11:35:32Z
8 pages, 2 figures, 9 tables
null
null
PP-LCNet: A Lightweight CPU Convolutional Neural Network
['Cheng Cui', 'Tingquan Gao', 'Shengyun Wei', 'Yuning Du', 'Ruoyu Guo', 'Shuilong Dong', 'Bin Lu', 'Ying Zhou', 'X. Lv', 'Qiwen Liu', 'Xiaoguang Hu', 'Dianhai Yu', 'Yanjun Ma']
2,021
arXiv.org
127
31
['Computer Science']
2,109.15107
CrossAug: A Contrastive Data Augmentation Method for Debiasing Fact Verification Models
['Minwoo Lee', 'Seungpil Won', 'Juae Kim', 'Hwanhee Lee', 'Cheoneum Park', 'Kyomin Jung']
['cs.CL', 'cs.AI']
Fact verification datasets are typically constructed using crowdsourcing techniques due to the lack of text sources with veracity labels. However, the crowdsourcing process often produces undesired biases in data that cause models to learn spurious patterns. In this paper, we propose CrossAug, a contrastive data augmen...
2021-09-30T13:19:19Z
5 pages, accepted as a short paper at CIKM 2021
null
10.1145/3459637.3482078
null
null
null
null
null
null
null
2,109.15254
SlovakBERT: Slovak Masked Language Model
['Matúš Pikuliak', 'Štefan Grivalský', 'Martin Konôpka', 'Miroslav Blšták', 'Martin Tamajka', 'Viktor Bachratý', 'Marián Šimko', 'Pavol Balážik', 'Michal Trnka', 'Filip Uhlárik']
['cs.CL']
We introduce a new Slovak masked language model called SlovakBERT. This is to our best knowledge the first paper discussing Slovak transformers-based language models. We evaluate our model on several NLP tasks and achieve state-of-the-art results. This evaluation is likewise the first attempt to establish a benchmark f...
2021-09-30T16:36:49Z
12 pages, 2 figures
null
null
SlovakBERT: Slovak Masked Language Model
['Matúš Pikuliak', 'Stefan Grivalsky', 'Martin Konopka', 'Miroslav Blšták', 'Martin Tamajka', "Viktor Bachrat'y", 'Marián Simko', 'Pavol Balázik', 'Michal Trnka', "Filip Uhl'arik"]
2,021
Conference on Empirical Methods in Natural Language Processing
27
49
['Computer Science']
2,110.00061
PubTables-1M: Towards comprehensive table extraction from unstructured documents
['Brandon Smock', 'Rohith Pesala', 'Robin Abraham']
['cs.LG', 'cs.CV']
Recently, significant progress has been made applying machine learning to the problem of table structure inference and extraction from unstructured documents. However, one of the greatest challenges remains the creation of datasets with complete, unambiguous ground truth at scale. To address this, we develop a new, mor...
2021-09-30T19:42:07Z
null
null
null
PubTables-1M: Towards comprehensive table extraction from unstructured documents
['B. Smock', 'Rohith Pesala', 'Robin Abraham']
2,021
Computer Vision and Pattern Recognition
103
28
['Computer Science']
2,110.00075
Noise2Recon: Enabling Joint MRI Reconstruction and Denoising with Semi-Supervised and Self-Supervised Learning
['Arjun D Desai', 'Batu M Ozturkler', 'Christopher M Sandino', 'Robert Boutin', 'Marc Willis', 'Shreyas Vasanawala', 'Brian A Hargreaves', 'Christopher M Ré', 'John M Pauly', 'Akshay S Chaudhari']
['eess.IV', 'cs.CV']
Deep learning (DL) has shown promise for faster, high quality accelerated MRI reconstruction. However, supervised DL methods depend on extensive amounts of fully-sampled (labeled) data and are sensitive to out-of-distribution (OOD) shifts, particularly low signal-to-noise ratio (SNR) acquisitions. To alleviate this cha...
2021-09-30T20:06:43Z
null
null
null
Noise2Recon: Enabling Joint MRI Reconstruction and Denoising with Semi-Supervised and Self-Supervised Learning
['Arjun D Desai', 'Batu Mehmet Ozturkler', 'Christopher M. Sandino', 'R. Boutin', 'M. Willis', 'S. Vasanawala', 'B. Hargreaves', 'Christopher Ré', 'J. Pauly', 'A. Chaudhari']
2,021
null
3
72
['Engineering', 'Computer Science']
2,110.00476
ResNet strikes back: An improved training procedure in timm
['Ross Wightman', 'Hugo Touvron', 'Hervé Jégou']
['cs.CV', 'cs.LG']
The influential Residual Networks designed by He et al. remain the gold-standard architecture in numerous scientific publications. They typically serve as the default architecture in studies, or as baselines when new architectures are proposed. Yet there has been significant progress on best practices for training neur...
2021-10-01T15:09:22Z
null
null
null
null
null
null
null
null
null
null
2,110.00976
LexGLUE: A Benchmark Dataset for Legal Language Understanding in English
['Ilias Chalkidis', 'Abhik Jana', 'Dirk Hartung', 'Michael Bommarito', 'Ion Androutsopoulos', 'Daniel Martin Katz', 'Nikolaos Aletras']
['cs.CL']
Laws and their interpretations, legal arguments and agreements\ are typically expressed in writing, leading to the production of vast corpora of legal text. Their analysis, which is at the center of legal practice, becomes increasingly elaborate as these collections grow in size. Natural language understanding (NLU) te...
2021-10-03T10:50:51Z
9 pages, long paper at ACL 2022 proceedings. LexGLUE benchmark is available at: https://huggingface.co/datasets/lex_glue. Code is available at: https://github.com/coastalcph/lex-glue. Update TFIDF-SVM scores in the last version
null
null
null
null
null
null
null
null
null
2,110.01485
JuriBERT: A Masked-Language Model Adaptation for French Legal Text
['Stella Douka', 'Hadi Abdine', 'Michalis Vazirgiannis', 'Rajaa El Hamdani', 'David Restrepo Amariles']
['cs.CL']
Language models have proven to be very useful when adapted to specific domains. Nonetheless, little research has been done on the adaptation of domain-specific BERT models in the French language. In this paper, we focus on creating a language model adapted to French legal text with the goal of helping law professionals...
2021-10-04T14:51:24Z
7 pages
null
null
null
null
null
null
null
null
null
2,110.01509
DeepA2: A Modular Framework for Deep Argument Analysis with Pretrained Neural Text2Text Language Models
['Gregor Betz', 'Kyle Richardson']
['cs.CL', 'cs.AI']
In this paper, we present and implement a multi-dimensional, modular framework for performing deep argument analysis (DeepA2) using current pre-trained language models (PTLMs). ArgumentAnalyst -- a T5 model (Raffel et al. 2020) set up and trained within DeepA2 -- reconstructs argumentative texts, which advance an infor...
2021-10-04T15:24:07Z
A Demo is available at https://huggingface.co/spaces/debatelab/deepa2-demo , the model can be downloaded from https://huggingface.co/debatelab/argument-analyst , and the datasets can be accessed at https://huggingface.co/datasets/debatelab/aaac
*SEM 2022
null
DeepA2: A Modular Framework for Deep Argument Analysis with Pretrained Neural Text2Text Language Models
['Gregor Betz', 'Kyle Richardson']
2,021
STARSEM
8
73
['Computer Science']
2,110.01518
Generalization in NLI: Ways (Not) To Go Beyond Simple Heuristics
['Prajjwal Bhargava', 'Aleksandr Drozd', 'Anna Rogers']
['cs.CL']
Much of recent progress in NLU was shown to be due to models' learning dataset-specific heuristics. We conduct a case study of generalization in NLI (from MNLI to the adversarially constructed HANS dataset) in a range of BERT-based architectures (adapters, Siamese Transformers, HEX debiasing), as well as with subsampli...
2021-10-04T15:37:07Z
Workshop on Insights from Negative Results (EMNLP 2021)
null
null
null
null
null
null
null
null
null
2,110.0171
PyTorrent: A Python Library Corpus for Large-scale Language Models
['Mehdi Bahrami', 'N. C. Shrikanth', 'Shade Ruangwan', 'Lei Liu', 'Yuji Mizobuchi', 'Masahiro Fukuyori', 'Wei-Peng Chen', 'Kazuki Munakata', 'Tim Menzies']
['cs.SE']
A large scale collection of both semantic and natural language resources is essential to leverage active Software Engineering research areas such as code reuse and code comprehensibility. Existing machine learning models ingest data from Open Source repositories (like GitHub projects) and forum discussions (like Stacko...
2021-10-04T20:48:31Z
10 pages, 2 figures, 5 tables
null
null
PyTorrent: A Python Library Corpus for Large-scale Language Models
['M. Bahrami', 'Shrikanth N. C.', 'Shade Ruangwan', 'Lei Liu', 'Yuji Mizobuchi', 'M. Fukuyori', 'Wei-Peng Chen', 'Kazuki Munakata', 'T. Menzies']
2,021
arXiv.org
12
38
['Computer Science']
2,110.01786
MoEfication: Transformer Feed-forward Layers are Mixtures of Experts
['Zhengyan Zhang', 'Yankai Lin', 'Zhiyuan Liu', 'Peng Li', 'Maosong Sun', 'Jie Zhou']
['cs.CL']
Recent work has shown that feed-forward networks (FFNs) in pre-trained Transformers are a key component, storing various linguistic and factual knowledge. However, the computational patterns of FFNs are still unclear. In this work, we study the computational patterns of FFNs and observe that most inputs only activate a...
2021-10-05T02:14:38Z
Accepted to ACL Findings 2022
null
null
MoEfication: Transformer Feed-forward Layers are Mixtures of Experts
['Zhengyan Zhang', 'Yankai Lin', 'Zhiyuan Liu', 'Peng Li', 'Maosong Sun', 'Jie Zhou']
2,021
Findings
129
67
['Computer Science']
2,110.01799
ContractNLI: A Dataset for Document-level Natural Language Inference for Contracts
['Yuta Koreeda', 'Christopher D. Manning']
['cs.CL', 'cs.AI', 'cs.LG']
Reviewing contracts is a time-consuming procedure that incurs large expenses to companies and social inequality to those who cannot afford it. In this work, we propose "document-level natural language inference (NLI) for contracts", a novel, real-world application of NLI that addresses such problems. In this task, a sy...
2021-10-05T03:22:31Z
Accepted at the Findings of the Association for Computational Linguistics: EMNLP 2021
null
null
ContractNLI: A Dataset for Document-level Natural Language Inference for Contracts
['Yuta Koreeda', 'Christopher D. Manning']
2,021
Conference on Empirical Methods in Natural Language Processing
106
24
['Computer Science']
2,110.019
DistilHuBERT: Speech Representation Learning by Layer-wise Distillation of Hidden-unit BERT
['Heng-Jui Chang', 'Shu-wen Yang', 'Hung-yi Lee']
['cs.CL', 'eess.AS']
Self-supervised speech representation learning methods like wav2vec 2.0 and Hidden-unit BERT (HuBERT) leverage unlabeled speech data for pre-training and offer good representations for numerous speech processing tasks. Despite the success of these methods, they require large memory and high pre-training costs, making t...
2021-10-05T09:34:44Z
Accepted to ICASSP 2022
null
null
Distilhubert: Speech Representation Learning by Layer-Wise Distillation of Hidden-Unit Bert
['Heng-Jui Chang', 'Shu-Wen Yang', 'Hung-yi Lee']
2,021
IEEE International Conference on Acoustics, Speech, and Signal Processing
175
36
['Computer Science', 'Engineering']
2,110.01938
Sicilian Translator: A Recipe for Low-Resource NMT
['Eryk Wdowiak']
['cs.CL', 'I.2.7']
With 17,000 pairs of Sicilian-English translated sentences, Arba Sicula developed the first neural machine translator for the Sicilian language. Using small subword vocabularies, we trained small Transformer models with high dropout parameters and achieved BLEU scores in the upper 20s. Then we supplemented our dataset ...
2021-10-05T11:04:13Z
7 pages, 2 tables
null
null
null
null
null
null
null
null
null
2,110.0203
Exploiting Twitter as Source of Large Corpora of Weakly Similar Pairs for Semantic Sentence Embeddings
['Marco Di Giovanni', 'Marco Brambilla']
['cs.CL']
Semantic sentence embeddings are usually supervisedly built minimizing distances between pairs of embeddings of sentences labelled as semantically similar by annotators. Since big labelled datasets are rare, in particular for non-English languages, and expensive, recent studies focus on unsupervised approaches that req...
2021-10-05T13:21:40Z
9 pages, 3 figures, accepted at EMNLP2021
null
null
null
null
null
null
null
null
null
2,110.02178
MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer
['Sachin Mehta', 'Mohammad Rastegari']
['cs.CV', 'cs.AI', 'cs.LG']
Light-weight convolutional neural networks (CNNs) are the de-facto for mobile vision tasks. Their spatial inductive biases allow them to learn representations with fewer parameters across different vision tasks. However, these networks are spatially local. To learn global representations, self-attention-based vision tr...
2021-10-05T17:07:53Z
Accepted at ICLR'22
null
null
MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer
['Sachin Mehta', 'Mohammad Rastegari']
2,021
International Conference on Learning Representations
1,306
65
['Computer Science']
2,110.02442
PoNet: Pooling Network for Efficient Token Mixing in Long Sequences
['Chao-Hong Tan', 'Qian Chen', 'Wen Wang', 'Qinglin Zhang', 'Siqi Zheng', 'Zhen-Hua Ling']
['cs.CL', 'cs.AI', 'cs.LG']
Transformer-based models have achieved great success in various NLP, vision, and speech tasks. However, the core of Transformer, the self-attention mechanism, has a quadratic time and memory complexity with respect to the sequence length, which hinders applications of Transformer-based models to long sequences. Many ap...
2021-10-06T01:07:54Z
Accepted by ICLR 2022. Codes and checkpoints are also available on huggingface hub: https://huggingface.co/chtan/ponet-base-uncased
null
null
null
null
null
null
null
null
null
2,110.02711
DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation
['Gwanghyun Kim', 'Taesung Kwon', 'Jong Chul Ye']
['cs.CV', 'cs.AI', 'cs.LG']
Recently, GAN inversion methods combined with Contrastive Language-Image Pretraining (CLIP) enables zero-shot image manipulation guided by text prompts. However, their applications to diverse real images are still difficult due to the limited GAN inversion capability. Specifically, these approaches often have difficult...
2021-10-06T12:59:39Z
Accepted to CVPR 2022
null
null
DiffusionCLIP: Text-Guided Diffusion Models for Robust Image Manipulation
['Gwanghyun Kim', 'Taesung Kwon', 'Jong-Chul Ye']
2,021
Computer Vision and Pattern Recognition
657
69
['Computer Science']
2,110.02861
8-bit Optimizers via Block-wise Quantization
['Tim Dettmers', 'Mike Lewis', 'Sam Shleifer', 'Luke Zettlemoyer']
['cs.LG']
Stateful optimizers maintain gradient statistics over time, e.g., the exponentially smoothed sum (SGD with momentum) or squared sum (Adam) of past gradient values. This state can be used to accelerate optimization compared to plain stochastic gradient descent but uses memory that might otherwise be allocated to model p...
2021-10-06T15:43:20Z
ICLR2022 spotlight version
null
null
null
null
null
null
null
null
null
2,110.03546
mRAT-SQL+GAP:A Portuguese Text-to-SQL Transformer
['Marcelo Archanjo José', 'Fabio Gagliardi Cozman']
['cs.CL', 'cs.AI', '68T07, 68T50', 'I.2.7; H.3.3']
The translation of natural language questions to SQL queries has attracted growing attention, in particular in connection with transformers and similar language models. A large number of techniques are geared towards the English language; in this work, we thus investigated translation to SQL when input questions are gi...
2021-10-07T15:08:24Z
Published in: Intelligent Systems. BRACIS 2021. Lecture Notes in Computer Science
vol 13074, 2021, pp 511-525
10.1007/978-3-030-91699-2_35
null
null
null
null
null
null
null
2,110.03584
Mixer-TTS: non-autoregressive, fast and compact text-to-speech model conditioned on language model embeddings
['Oktai Tatanov', 'Stanislav Beliaev', 'Boris Ginsburg']
['eess.AS']
This paper describes Mixer-TTS, a non-autoregressive model for mel-spectrogram generation. The model is based on the MLP-Mixer architecture adapted for speech synthesis. The basic Mixer-TTS contains pitch and duration predictors, with the latter being trained with an unsupervised TTS alignment framework. Alongside the ...
2021-10-07T16:07:58Z
Preprint. Submitted to ICASSP-22
null
null
Mixer-TTS: Non-Autoregressive, Fast and Compact Text-to-Speech Model Conditioned on Language Model Embeddings
['Oktai Tatanov', 'Stanislav Beliaev', 'Boris Ginsburg']
2,021
IEEE International Conference on Acoustics, Speech, and Signal Processing
16
23
['Engineering', 'Computer Science']
2,110.03895
ALL-IN-ONE: Multi-Task Learning BERT models for Evaluating Peer Assessments
['Qinjin Jia', 'Jialin Cui', 'Yunkai Xiao', 'Chengyuan Liu', 'Parvez Rashid', 'Edward F. Gehringer']
['cs.CL', 'cs.AI']
Peer assessment has been widely applied across diverse academic fields over the last few decades and has demonstrated its effectiveness. However, the advantages of peer assessment can only be achieved with high-quality peer reviews. Previous studies have found that high-quality review comments usually comprise several ...
2021-10-08T05:13:41Z
null
null
null
null
null
null
null
null
null
null
2,110.04057
FAST-RIR: Fast neural diffuse room impulse response generator
['Anton Ratnarajah', 'Shi-Xiong Zhang', 'Meng Yu', 'Zhenyu Tang', 'Dinesh Manocha', 'Dong Yu']
['cs.SD', 'cs.AI', 'cs.LG', 'eess.AS']
We present a neural-network-based fast diffuse room impulse response generator (FAST-RIR) for generating room impulse responses (RIRs) for a given acoustic environment. Our FAST-RIR takes rectangular room dimensions, listener and speaker positions, and reverberation time as inputs and generates specular and diffuse ref...
2021-10-07T05:21:01Z
Accepted to ICASSP 2022. More results and source code is available at https://anton-jeran.github.io/FRIR/
null
null
null
null
null
null
null
null
null
2,110.0441
TitaNet: Neural Model for speaker representation with 1D Depth-wise separable convolutions and global context
['Nithin Rao Koluguri', 'Taejin Park', 'Boris Ginsburg']
['eess.AS', 'cs.SD']
In this paper, we propose TitaNet, a novel neural network architecture for extracting speaker representations. We employ 1D depth-wise separable convolutions with Squeeze-and-Excitation (SE) layers with global context followed by channel attention based statistics pooling layer to map variable-length utterances to a fi...
2021-10-08T23:49:42Z
preprint. Submitted to ICASSP 2022
null
null
TitaNet: Neural Model for Speaker Representation with 1D Depth-Wise Separable Convolutions and Global Context
['N. Koluguri', 'Taejin Park', 'Boris Ginsburg']
2,021
IEEE International Conference on Acoustics, Speech, and Signal Processing
104
32
['Computer Science', 'Engineering']
2,110.04725
Yuan 1.0: Large-Scale Pre-trained Language Model in Zero-Shot and Few-Shot Learning
['Shaohua Wu', 'Xudong Zhao', 'Tong Yu', 'Rongguo Zhang', 'Chong Shen', 'Hongli Liu', 'Feng Li', 'Hong Zhu', 'Jiangang Luo', 'Liang Xu', 'Xuanwei Zhang']
['cs.CL', 'cs.AI']
Recent work like GPT-3 has demonstrated excellent performance of Zero-Shot and Few-Shot learning on many natural language processing (NLP) tasks by scaling up model size, dataset size and the amount of computation. However, training a model like GPT-3 requires huge amount of computational resources which makes it chall...
2021-10-10T07:40:22Z
null
null
null
null
null
null
null
null
null
null
2,110.04994
Omnidata: A Scalable Pipeline for Making Multi-Task Mid-Level Vision Datasets from 3D Scans
['Ainaz Eftekhar', 'Alexander Sax', 'Roman Bachmann', 'Jitendra Malik', 'Amir Zamir']
['cs.CV', 'cs.AI', 'cs.GR', 'cs.RO']
This paper introduces a pipeline to parametrically sample and render multi-task vision datasets from comprehensive 3D scans from the real world. Changing the sampling parameters allows one to "steer" the generated datasets to emphasize specific information. In addition to enabling interesting lines of research, we show...
2021-10-11T04:21:46Z
ICCV 2021: See project website https://omnidata.vision
null
null
null
null
null
null
null
null
null
2,110.05069
Efficient Training of Audio Transformers with Patchout
['Khaled Koutini', 'Jan Schlüter', 'Hamid Eghbal-zadeh', 'Gerhard Widmer']
['cs.SD', 'cs.LG', 'eess.AS']
The great success of transformer-based models in natural language processing (NLP) has led to various attempts at adapting these architectures to other domains such as vision and audio. Recent work has shown that transformers can outperform Convolutional Neural Networks (CNNs) on vision and audio tasks. However, one of...
2021-10-11T08:07:50Z
Submitted to Interspeech 2022. Source code: https://github.com/kkoutini/PaSST
null
10.21437/Interspeech.2022-227
Efficient Training of Audio Transformers with Patchout
['Khaled Koutini', 'Jan Schlüter', 'Hamid Eghbalzadeh', 'G. Widmer']
2,021
Interspeech
263
30
['Computer Science', 'Engineering']
2,110.05679
Large Language Models Can Be Strong Differentially Private Learners
['Xuechen Li', 'Florian Tramèr', 'Percy Liang', 'Tatsunori Hashimoto']
['cs.LG', 'cs.CL']
Differentially Private (DP) learning has seen limited success for building large deep learning models of text, and straightforward attempts at applying Differentially Private Stochastic Gradient Descent (DP-SGD) to NLP tasks have resulted in large performance drops and high computational overhead. We show that this per...
2021-10-12T01:45:27Z
31 pages; update ethics statement to clarify benefits and potential long-term harms
null
null
null
null
null
null
null
null
null
2,110.05752
UniSpeech-SAT: Universal Speech Representation Learning with Speaker Aware Pre-Training
['Sanyuan Chen', 'Yu Wu', 'Chengyi Wang', 'Zhengyang Chen', 'Zhuo Chen', 'Shujie Liu', 'Jian Wu', 'Yao Qian', 'Furu Wei', 'Jinyu Li', 'Xiangzhan Yu']
['cs.CL', 'cs.SD', 'eess.AS']
Self-supervised learning (SSL) is a long-standing goal for speech processing, since it utilizes large-scale unlabeled data and avoids extensive human labeling. Recent years witness great successes in applying self-supervised learning in speech recognition, while limited exploration was attempted in applying SSL for mod...
2021-10-12T05:43:30Z
ICASSP 2022 Submission
null
null
null
null
null
null
null
null
null
2,110.05781
BERTraffic: BERT-based Joint Speaker Role and Speaker Change Detection for Air Traffic Control Communications
['Juan Zuluaga-Gomez', 'Seyyed Saeed Sarfjoo', 'Amrutha Prasad', 'Iuliia Nigmatulina', 'Petr Motlicek', 'Karel Ondrej', 'Oliver Ohneiser', 'Hartmut Helmke']
['eess.AS', 'cs.CL', 'cs.LG']
Automatic speech recognition (ASR) allows transcribing the communications between air traffic controllers (ATCOs) and aircraft pilots. The transcriptions are used later to extract ATC named entities, e.g., aircraft callsigns. One common challenge is speech activity detection (SAD) and speaker diarization (SD). In the f...
2021-10-12T07:25:12Z
To be published in the 2022 IEEE Spoken Language Technology Workshop (SLT) (SLT 2022)
null
null
null
null
null
null
null
null
null