Filters








23,234 Hits in 8.3 sec

A speech interface for open-domain question-answering

Edward Schofield, Zhiping Zheng
2003 Proceedings of the 41st Annual Meeting on Association for Computational Linguistics - ACL '03  
robustness to spoken input.  ...  We describe a small evaluation of the effect of recognition errors on the precision of the answers returned and make some concrete recommendations for modifying a question-answering system for improving  ...  Acknowledgements The authors would like to thank Stefan Rüger for his suggestions and moral support. Ed Schofield's research is supported by a Marie Curie Fellowship of the European Commission.  ... 
doi:10.3115/1075178.1075210 dblp:conf/acl/SchofieldZ03 fatcat:hfbcjhjroraf7djzsurkctxc54

DUAL: Discrete Spoken Unit Adaptive Learning for Textless Spoken Question Answering [article]

Guan-Ting Lin, Yung-Sung Chuang, Ho-Lam Chung, Shu-wen Yang, Hsuan-Jui Chen, Shuyan Dong, Shang-Wen Li, Abdelrahman Mohamed, Hung-yi Lee, Lin-shan Lee
2022 arXiv   pre-print
Spoken Question Answering (SQA) is to find the answer from a spoken document given a question, which is crucial for personal assistants when replying to the queries from the users.  ...  Also, ASR aims to minimize recognition errors equally over all words, including many function words irrelevant to the SQA task.  ...  , including adapting the TQA model to be more robust against recognition errors.  ... 
arXiv:2203.04911v3 fatcat:jthifjtzebdf5gt6w22chy3hxa

Talk to Me: Verbal Communication Improves Perceptions of Friendship and Social Presence in Human-Robot Interaction [chapter]

Elena Corina Grigore, Andre Pereira, Ian Zhou, David Wang, Brian Scassellati
2016 Lecture Notes in Computer Science  
Our results highlight the advantages of using speech recognition as an input communication modality even when this represents the less robust choice, and the importance of investigating how to best do  ...  In this work, we consider how two core input communication modalities present in human-robot interaction-speech recognition and touch-based selection-shape users' perceptions of the agent.  ...  Users in this condition would say one of the options displayed on the screen to answer a question. We employed the Google Speech Recognition API [12].  ... 
doi:10.1007/978-3-319-47665-0_5 fatcat:evd3p6bt3refnmd7hulrjgm4vq

Effects of Language Modeling on Speech-driven Question Answering [article]

Tomoyosi Akiba, Atsushi Fujii, Katunobu Itou
2004 arXiv   pre-print
We propose a passage retrieval method robust against recognition errors in the transcription.  ...  We integrate automatic speech recognition (ASR) and question answering (QA) to realize a speech-driven QA system, and evaluate its performance.  ...  Because by definition speech-driven question answering accepts a result of speech recognition as an input, which often includes errors, the passage retrieval must be robust against those errors.  ... 
arXiv:cs/0407028v1 fatcat:quv7dmt4ebhtzhjtifxtgpcgem

Integration of speech with natural language understanding

R. C. Moore
1995 Proceedings of the National Academy of Sciences of the United States of America  
Finally, I discuss how prosodic information in the speech signal might be used to improve understanding.  ...  I then look at how systems cope with errors in speech recognition and at attempts to use natural language information to reduce recognition errors.  ...  ROBUSTNESS TO RECOGNITION ERRORS Since even the best speech recognition systems make at least some errors in a substantial proportion of utterances, coping with speech recognition errors is one of the  ... 
doi:10.1073/pnas.92.22.9983 pmid:7479813 pmcid:PMC40722 fatcat:iyynjeqinjeyxjbnj2siqfjft4

Experiments in speech driven question answering

C. Gonzalez-Ferreras, V. Cardenoso-Payo, E. Sanchis Arnal
2008 2008 IEEE Spoken Language Technology Workshop  
Some improvements over the baseline system were proposed in order to adapt the output of the speech recognizer to the question answering engine: capitalized output from the speech recognizer and a language  ...  A large vocabulary continuous speech recognizer is used to transcribe the spoken question into text. Then, a question answering engine is used to obtain the answer to the question.  ...  In this work we present an approach to speech driven question answering in which we study the problem of adapting the output of the speech recognizer to the question answering engine.  ... 
doi:10.1109/slt.2008.4777846 dblp:conf/slt/FerrerasCA08 fatcat:dcwnuh3rqrgk7h4ii3tzzm6d7u

ODSQA: Open-domain Spoken Question Answering Dataset [article]

Chia-Hsuan Lee and Shang-Ming Wang and Huan-Cheng Chang and Hung-Yi Lee
2018 arXiv   pre-print
In this paper, we release Open-Domain Spoken Question Answering Dataset (ODSQA) with more than three thousand questions. To the best of our knowledge, this is the largest real SQA dataset.  ...  On this dataset, we found that ASR errors have catastrophic impact on SQA. To mitigate the effect of ASR errors, subword units are involved, which brings consistent improvements over all the models.  ...  To improve the robustness to speech recognition errors, we used back-translation as a data augmentation approach in this paper.  ... 
arXiv:1808.02280v1 fatcat:bu4i3cbha5edto346fr2mews7a

On the Voice-Activated Question Answering

Paolo Rosso, Lluís-F. Hurtado, Encarna Segarra, Emilio Sanchis
2012 IEEE Transactions on Systems Man and Cybernetics Part C (Applications and Reviews)  
Finally, the scenarios that will emerge from the introduction of speech recognition in Question Answering will be discussed.  ...  Abstract Question Answering is probably one of the most challenging tasks in the field of Natural Language Processing.  ...  The preliminary experiments with simulated speech input (i.e., errors in the input questions -200 questions of the CLEF 2005 Finally , we would like to point out the interest of Information Retrieval  ... 
doi:10.1109/tsmcc.2010.2089620 fatcat:tirmodeurbg7rkok6juj7zdet4

The roles of language processing in a spoken language interface

L. Hirschman
1995 Proceedings of the National Academy of Sciences of the United States of America  
The paper reviews the dual role of language processing in providing undersnding of the spoken input and an additional source ofconstmint in the recognition process. To date, language processing has  ...  This paper provides an overview of the colloquium's discussion session on natural language understanding, which followed presentations by M. Bates [Bates, M. (1995) Proc. Natl. Acad. Sci.  ...  input and how much is due to recognition errors.  ... 
doi:10.1073/pnas.92.22.9970 pmid:7479811 pmcid:PMC40720 fatcat:qlaxn6zs45gr3mc76xfvpppaeu

A Robust System for Natural Spoken Dialogue [article]

James F. Allen, Bradford W. Miller, Eric K. Ringger, Teresa Sikorski
1996 arXiv   pre-print
It specifically addresses the issue of robust interpretation of speech in the presence of recognition errors.  ...  This paper describes a system that leads us to believe in the feasibility of constructing natural spoken dialogue systems in task-oriented domains.  ...  Many thanks to Alex Rudnicky, Ronald Rosenfeld and Sunil Issar at CMU for providing the Sphinx-II system and related tools.  ... 
arXiv:cmp-lg/9606023v1 fatcat:tahoou6wircxnooyj57ch67g7m

Atypical Speech

Georg Stemmer, Elmar Nöth, Vijay Parsa
2010 EURASIP Journal on Audio, Speech, and Music Processing  
We hope that the papers in this special issue help to advance in the direction of getting an answer to these questions.  ...  The majority of the articles analyses the influence of atypical speech on automatic speech recognition performance in great detail, and different methods to reduce the error rates for atypical speech are  ...  We hope that the papers in this special issue help to advance in the direction of getting an answer to these questions.  ... 
doi:10.1155/2010/835974 fatcat:trnubdf6k5ditchy3grxg7obom

A robust system for natural spoken dialogue

James F. Allen, Bradford W. Miller, Eric K. Ringger, Teresa Sikorski
1996 Proceedings of the 34th annual meeting on Association for Computational Linguistics -  
It specifically addresses the issue of robust interpretation of speech in the presence of recognition errors.  ...  This paper describes a system that leads us to believe in the feasibility of constructing natural spoken dialogue systems in task-oriented domains.  ...  Many thanks to Alex Rudnicky, Ronald Rosenfeld and Sunil Issar at CMU for providing the Sphinx-II system and related tools.  ... 
doi:10.3115/981863.981872 dblp:conf/acl/AllenMRS96 fatcat:ue7dybxtazfghirxb6lw5obfwi

Dependency of recognition rate on number of words for text‐independent speaker recognition using vector quantization

Hidenori Shimizu, Tetsuo Funada
2008 Journal of the Acoustical Society of America  
According to speaker identification experiments, increasing the number of words in recognition more than three words, this system can attain a correct rate of 100% for input speech of 40 speakers.  ...  The feature FTTSS is used to develop a robust speaker recognition system under noisy condition.  ...  The case where the person in question is correctly identified is classified to be a correct answer.  ... 
doi:10.1121/1.2935784 fatcat:g6sjywcah5dtdift3hvdtujhf4

Self-supervised Contrastive Cross-Modality Representation Learning for Spoken Question Answering [article]

Chenyu You, Nuo Chen, Yuexian Zou
2021 arXiv   pre-print
Besides, we design a Temporal-Alignment attention to semantically align the speech-text clues in the learned common space and benefit the SQA tasks.  ...  Spoken question answering (SQA) requires fine-grained understanding of both spoken documents and questions for the optimal answer prediction.  ...  and introduced sub-word unit strategy to alleviate the effects of speech recognition errors in SQA.  ... 
arXiv:2109.03381v1 fatcat:qdt3ufhby5ao7h2j6hj2nag4p4

Mitigating Noisy Inputs for Question Answering [article]

Denis Peskov, Joe Barrow, Pedro Rodriguez, Graham Neubig, Jordan Boyd-Graber
2019 arXiv   pre-print
We investigate and mitigate the effects of noise from Automatic Speech Recognition systems on two factoid Question Answering (QA) tasks.  ...  Integrating confidences into the model and forced decoding of unknown words are empirically shown to improve the accuracy of downstream neural QA systems.  ...  However, most factoid QA models are trained and evaluated on clean text input, which becomes noisy when questions are spoken due to Automatic Speech Recognition (ASR) errors.  ... 
arXiv:1908.02914v1 fatcat:4mk2tlcuvbej5ly5ngniuivgwy
« Previous Showing results 1 — 15 out of 23,234 results