Filters








6,193 Hits in 6.1 sec

Data Augmentation for Spoken Language Understanding via Joint Variational Generation [article]

Kang Min Yoo, Youhyun Shin, Sang-goo Lee
2018 arXiv   pre-print
Data scarcity is one of the main obstacles of domain adaptation in spoken language understanding (SLU) due to the high cost of creating manually tagged SLU datasets.  ...  Our approach not only helps alleviate the data scarcity issue in the SLU task for many datasets but also indiscriminately improves language understanding performances for various SLU models, supported  ...  However, Figure 1 : The general framework for generative language understanding data augmentation.  ... 
arXiv:1809.02305v2 fatcat:lrab2b45yjbwhgwxjyytqdeqxe

Data Augmentation for Spoken Language Understanding via Joint Variational Generation

Kang Min Yoo, Youhyun Shin, Sang-goo Lee
2019 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Data scarcity is one of the main obstacles of domain adaptation in spoken language understanding (SLU) due to the high cost of creating manually tagged SLU datasets.  ...  Our approach not only helps alleviate the data scarcity issue in the SLU task for many datasets but also indiscriminately improves language understanding performances for various SLU models, supported  ...  from our proposed generative model, Joint Language Understanding Variational Autoencoder (JLUVA).  ... 
doi:10.1609/aaai.v33i01.33017402 fatcat:36woenu7s5ds5n5wtl4kw44ovu

Continuous 3D Multi-Channel Sign Language Production via Progressive Transformers and Mixture Density Networks [article]

Ben Saunders, Necati Cihan Camgoz, Richard Bowden
2021 arXiv   pre-print
Sign languages are multi-channel visual languages, where signers use a continuous 3D space to communicate.Sign Language Production (SLP), the automatic translation from spoken to sign languages, must embody  ...  We present extensive data augmentation techniques to reduce prediction drift, alongside an adversarial training regime and a Mixture Density Network (MDN) formulation to produce realistic and expressive  ...  Acknowledgements We would like to thank Tao Jiang for their help with data curation.  ... 
arXiv:2103.06982v1 fatcat:kqgcjywhojgapjbvguf73pqvfu

Continuous 3D Multi-Channel Sign Language Production via Progressive Transformers and Mixture Density Networks

Ben Saunders, Necati Cihan Camgoz, Richard Bowden
2021 International Journal of Computer Vision  
Sign language production (SLP), the automatic translation from spoken to sign languages, must embody both the continuous articulation and full morphology of sign to be truly understandable by the Deaf  ...  We present extensive data augmentation techniques to reduce prediction drift, alongside an adversarial training regime and a mixture density network (MDN) formulation to produce realistic and expressive  ...  This avoids the large variation in small joint positions a large sigma would create, particularly for the hands.  ... 
doi:10.1007/s11263-021-01457-9 fatcat:d4um3whju5e25a45ctjjubxoie

Data Augmentation with Paraphrase Generation and Entity Extraction for Multimodal Dialogue System [article]

Eda Okur, Saurav Sahay, Lama Nachman
2022 arXiv   pre-print
This work explores the potential benefits of data augmentation with paraphrase generation for the NLU models trained on small task-specific datasets.  ...  Our focus is on improving the Natural Language Understanding (NLU) module of the task-oriented SDS pipeline with limited datasets.  ...  for our use-cases.  ... 
arXiv:2205.04006v1 fatcat:folvts7gjzgetjxqpppnimwzw4

Variational Hierarchical Dialog Autoencoder for Dialog State Tracking Data Augmentation [article]

Kang Min Yoo, Hanbit Lee, Franck Dernoncourt, Trung Bui, Walter Chang, Sang-goo Lee
2020 arXiv   pre-print
Experiments on various dialog datasets show that our model improves the downstream dialog trackers' robustness via generative data augmentation.  ...  Recent works have shown that generative data augmentation, where synthetic samples generated from deep generative models complement the training dataset, benefit NLP tasks.  ...  Acknowledgement We thank Hyunsoo Cho for his help with implementations and Jihun Choi for the thoughtful feedback.  ... 
arXiv:2001.08604v3 fatcat:ed4hldwkcnb2boihzhznqrm2t4

Data Augmentation with Atomic Templates for Spoken Language Understanding

Zijian Zhao, Su Zhu, Kai Yu
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
Spoken Language Understanding (SLU) converts user utterances into structured semantic representations.  ...  In this work, we propose a data augmentation method with atomic templates for SLU, which involves minimum human efforts.  ...  We thank the anonymous reviewers for their thoughtful comments and efforts towards improving this manuscript.  ... 
doi:10.18653/v1/d19-1375 dblp:conf/emnlp/ZhaoZY19 fatcat:7ldwfvq2azcnjaxqtagfp67pqm

Adapting to the Long Tail: A Meta-Analysis of Transfer Learning Research for Language Understanding Tasks [article]

Aakanksha Naik, Jill Lehman, Carolyn Rose
2021 arXiv   pre-print
Natural language understanding (NLU) has made massive progress driven by large benchmarks, paired with research on transfer learning to broaden its impact.  ...  Our answers to these questions highlight major avenues for future research in transfer learning for the long tail.  ...  Introduction "There is a growing consensus that significant, rapid progress can be made in both text understanding and spoken language understanding by investigating those phenomena that occur most centrally  ... 
arXiv:2111.01340v1 fatcat:rlg77auu3zfwdggxy3kwm7fu2m

Robustness Testing of Language Understanding in Task-Oriented Dialog [article]

Jiexi Liu, Ryuichi Takanobu, Jiaxin Wen, Dazhen Wan, Hongguang Li, Weiran Nie, Cheng Li, Wei Peng, Minlie Huang
2021 arXiv   pre-print
The augmented dataset through LAUG can be used to facilitate future research on the robustness testing of language understanding in task-oriented dialog.  ...  Most language understanding models in task-oriented dialog systems are trained on a small amount of annotated training data, and evaluated in a small set from the same distribution.  ...  We would like to thank colleagues from HUAWEI for their constant support and valuable discussion.  ... 
arXiv:2012.15262v3 fatcat:ymrcjqfunfcidn3uwcpo5rum5a

A Russian Keyword Spotting System Based on Large Vocabulary Continuous Speech Recognition and Linguistic Knowledge

Valentin Smirnov, Dmitry Ignatov, Michael Gusev, Mais Farkhadov, Natalia Rumyantseva, Mukhabbat Farkhadova
2016 Journal of Electrical and Computer Engineering  
Key algorithms and system settings are described, including the pronunciation variation algorithm, and the experimental results on the real-life telecom data are provided.  ...  The paper describes the key concepts of a word spotting system for Russian based on large vocabulary continuous speech recognition.  ...  Acknowledgments The authors would like to thank SpRecord LLC authorities for providing real-world telephone-quality data used in training and testing of the keyword spotting system described in this paper  ... 
doi:10.1155/2016/4062786 fatcat:7jhohy6kerbuln7drrwcqfizcq

Composed Variational Natural Language Generation for Few-shot Intents [article]

Congying Xia, Caiming Xiong, Philip Yu, Richard Socher
2020 arXiv   pre-print
To build connections between existing many-shot intents and few-shot intents, we consider an intent as a combination of a domain and an action, and propose a composed variational natural language generator  ...  In this paper, we focus on generating training examples for few-shot intents in the realistic imbalanced scenario.  ...  Acknowledgments We thank the reviewers for their valuable comments. This work is supported in part by NSF under grants III-1763325, III-1909323, and SaTC-1930941.  ... 
arXiv:2009.10056v1 fatcat:japh2i52zffg7pmko67k522rwq

Detailed author index

2009 2009 IEEE Workshop on Automatic Speech Recognition & Understanding  
Grounding of Spoken Language Understanding 479 The Exploration/Exploitation Trade-Off in Reinforcement Learning for Dialogue Management Rigoll, Gerhard 349 Robust Vocabulary Independent Keyword  ...  Spoken Keyword Spotting via Segmental DTW on Gaussian Posteriorgrams Zhou, Bowen 136 Improving Online Incremental Speaker Adaptation with Eigen Feature Space MLLR 502 Reinforcing Language Model for Speech  ... 
doi:10.1109/asru.2009.5373491 fatcat:tgoktcyotncatbvo3l3cslfapa

NoiseQA: Challenge Set Evaluation for User-Centric Question Answering [article]

Abhilasha Ravichander, Siddharth Dalmia, Maria Ryskina, Florian Metze, Eduard Hovy, Alan W Black
2021 arXiv   pre-print
We conclude that there is substantial room for progress before QA systems can be effectively deployed, highlight the need for QA evaluation to expand to consider real-world use, and hope that our findings  ...  systems are deployed in the real world, users query them through a variety of interfaces, such as speaking to voice assistants, typing questions into a search engine, or even translating questions to languages  ...  Acknowledgments We thank Aakanksha Naik, Sujeath Pareddy, Taylor Berg-Kirkpatrick, and Matthew Gormley for helpful discussion and the anonymous reviewers for their valuable feedback.  ... 
arXiv:2102.08345v1 fatcat:wwnvamh7krgxjctlipcojfb4qq

Robust Spoken Language Understanding via Paraphrasing [article]

Avik Ray, Yilin Shen, Hongxia Jin
2018 arXiv   pre-print
Learning intents and slot labels from user utterances is a fundamental step in all spoken language understanding (SLU) and dialog systems.  ...  We propose two new paraphrase generators using RNN and sequence-to-sequence based neural networks, which are suitable for our application.  ...  Spoken language understanding (SLU) unit, or a semantic parser lie at its core which enables the agent to map a user utterance to the corresponding action desired by the user.  ... 
arXiv:1809.06444v1 fatcat:c64ax2we7fgjvbi7pz7ndz6ram

Robust Spoken Language Understanding via Paraphrasing

Avik Ray, Yilin Shen, Hongxia Jin
2018 Interspeech 2018  
Learning intents and slot labels from user utterances is a fundamental step in all spoken language understanding (SLU) and dialog systems.  ...  We propose two new paraphrase generators using RNN and sequenceto-sequence based neural networks, which are suitable for our application.  ...  flights one way with first class arriving in tacoma Datasets: For evaluation we use the benchmark ATIS dataset [29] , which is popularly used for evaluating parsers for spoken language understanding.  ... 
doi:10.21437/interspeech.2018-2358 dblp:conf/interspeech/RaySJ18 fatcat:bhwzxuoglredppa5usvjb72fci
« Previous Showing results 1 — 15 out of 6,193 results