Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Tara N. Sainath

Tara N. Sainath

Lookup-Table Recurrent Language Models for Long Tail Speech Recognition


Apr 09, 2021
W. Ronny Huang, Tara N. Sainath, Cal Peyser, Shankar Kumar, David Rybach, Trevor Strohman

* Submitted to Interspeech 2021 
  Access Paper or Ask Questions

Learning Word-Level Confidence For Subword End-to-End ASR


Mar 11, 2021
David Qiu, Qiujia Li, Yanzhang He, Yu Zhang, Bo Li, Liangliang Cao, Rohit Prabhavalkar, Deepti Bhatia, Wei Li, Ke Hu, Tara N. Sainath, Ian McGraw

Figure 1 for Learning Word-Level Confidence For Subword End-to-End ASR
Figure 2 for Learning Word-Level Confidence For Subword End-to-End ASR
Figure 3 for Learning Word-Level Confidence For Subword End-to-End ASR
Figure 4 for Learning Word-Level Confidence For Subword End-to-End ASR
* To appear in ICASSP 2021 
  Access Paper or Ask Questions

Transformer Based Deliberation for Two-Pass Speech Recognition


Jan 27, 2021
Ke Hu, Ruoming Pang, Tara N. Sainath, Trevor Strohman

Figure 1 for Transformer Based Deliberation for Two-Pass Speech Recognition
Figure 2 for Transformer Based Deliberation for Two-Pass Speech Recognition
Figure 3 for Transformer Based Deliberation for Two-Pass Speech Recognition
Figure 4 for Transformer Based Deliberation for Two-Pass Speech Recognition

  Access Paper or Ask Questions

Less Is More: Improved RNN-T Decoding Using Limited Label Context and Path Merging


Dec 12, 2020
Rohit Prabhavalkar, Yanzhang He, David Rybach, Sean Campbell, Arun Narayanan, Trevor Strohman, Tara N. Sainath

Figure 1 for Less Is More: Improved RNN-T Decoding Using Limited Label Context and Path Merging
Figure 2 for Less Is More: Improved RNN-T Decoding Using Limited Label Context and Path Merging
Figure 3 for Less Is More: Improved RNN-T Decoding Using Limited Label Context and Path Merging
Figure 4 for Less Is More: Improved RNN-T Decoding Using Limited Label Context and Path Merging

  Access Paper or Ask Questions

Cascaded encoders for unifying streaming and non-streaming ASR


Oct 27, 2020
Arun Narayanan, Tara N. Sainath, Ruoming Pang, Jiahui Yu, Chung-Cheng Chiu, Rohit Prabhavalkar, Ehsan Variani, Trevor Strohman

Figure 1 for Cascaded encoders for unifying streaming and non-streaming ASR
Figure 2 for Cascaded encoders for unifying streaming and non-streaming ASR
Figure 3 for Cascaded encoders for unifying streaming and non-streaming ASR
Figure 4 for Cascaded encoders for unifying streaming and non-streaming ASR

  Access Paper or Ask Questions

Multitask Training with Text Data for End-to-End Speech Recognition


Oct 27, 2020
Peidong Wang, Tara N. Sainath, Ron J. Weiss

Figure 1 for Multitask Training with Text Data for End-to-End Speech Recognition
Figure 2 for Multitask Training with Text Data for End-to-End Speech Recognition
Figure 3 for Multitask Training with Text Data for End-to-End Speech Recognition
Figure 4 for Multitask Training with Text Data for End-to-End Speech Recognition

  Access Paper or Ask Questions

FastEmit: Low-latency Streaming ASR with Sequence-level Emission Regularization


Oct 21, 2020
Jiahui Yu, Chung-Cheng Chiu, Bo Li, Shuo-yiin Chang, Tara N. Sainath, Yanzhang He, Arun Narayanan, Wei Han, Anmol Gulati, Yonghui Wu, Ruoming Pang

Figure 1 for FastEmit: Low-latency Streaming ASR with Sequence-level Emission Regularization
Figure 2 for FastEmit: Low-latency Streaming ASR with Sequence-level Emission Regularization
Figure 3 for FastEmit: Low-latency Streaming ASR with Sequence-level Emission Regularization
Figure 4 for FastEmit: Low-latency Streaming ASR with Sequence-level Emission Regularization
* tech report 
  Access Paper or Ask Questions

Universal ASR: Unify and Improve Streaming ASR with Full-context Modeling


Oct 12, 2020
Jiahui Yu, Wei Han, Anmol Gulati, Chung-Cheng Chiu, Bo Li, Tara N. Sainath, Yonghui Wu, Ruoming Pang

Figure 1 for Universal ASR: Unify and Improve Streaming ASR with Full-context Modeling
Figure 2 for Universal ASR: Unify and Improve Streaming ASR with Full-context Modeling
Figure 3 for Universal ASR: Unify and Improve Streaming ASR with Full-context Modeling
Figure 4 for Universal ASR: Unify and Improve Streaming ASR with Full-context Modeling
* tech report 
  Access Paper or Ask Questions

Improving Tail Performance of a Deliberation E2E ASR Model Using a Large Text Corpus


Aug 25, 2020
Cal Peyser, Sepand Mavandadi, Tara N. Sainath, James Apfel, Ruoming Pang, Shankar Kumar

Figure 1 for Improving Tail Performance of a Deliberation E2E ASR Model Using a Large Text Corpus
Figure 2 for Improving Tail Performance of a Deliberation E2E ASR Model Using a Large Text Corpus
Figure 3 for Improving Tail Performance of a Deliberation E2E ASR Model Using a Large Text Corpus
Figure 4 for Improving Tail Performance of a Deliberation E2E ASR Model Using a Large Text Corpus

  Access Paper or Ask Questions

Improving Proper Noun Recognition in End-to-End ASR By Customization of the MWER Loss Criterion


May 19, 2020
Cal Peyser, Tara N. Sainath, Golan Pundak

Figure 1 for Improving Proper Noun Recognition in End-to-End ASR By Customization of the MWER Loss Criterion
Figure 2 for Improving Proper Noun Recognition in End-to-End ASR By Customization of the MWER Loss Criterion
Figure 3 for Improving Proper Noun Recognition in End-to-End ASR By Customization of the MWER Loss Criterion
Figure 4 for Improving Proper Noun Recognition in End-to-End ASR By Customization of the MWER Loss Criterion

  Access Paper or Ask Questions

RNN-T Models Fail to Generalize to Out-of-Domain Audio: Causes and Solutions


May 17, 2020
Chung-Cheng Chiu, Arun Narayanan, Wei Han, Rohit Prabhavalkar, Yu Zhang, Navdeep Jaitly, Ruoming Pang, Tara N. Sainath, Patrick Nguyen, Liangliang Cao, Yonghui Wu

Figure 1 for RNN-T Models Fail to Generalize to Out-of-Domain Audio: Causes and Solutions
Figure 2 for RNN-T Models Fail to Generalize to Out-of-Domain Audio: Causes and Solutions
Figure 3 for RNN-T Models Fail to Generalize to Out-of-Domain Audio: Causes and Solutions
Figure 4 for RNN-T Models Fail to Generalize to Out-of-Domain Audio: Causes and Solutions
* Submitted to Interspeech 2020 
  Access Paper or Ask Questions

A Streaming On-Device End-to-End Model Surpassing Server-Side Conventional Model Quality and Latency


Mar 28, 2020
Tara N. Sainath, Yanzhang He, Bo Li, Arun Narayanan, Ruoming Pang, Antoine Bruguier, Shuo-yiin Chang, Wei Li, Raziel Alvarez, Zhifeng Chen, Chung-Cheng Chiu, David Garcia, Alex Gruenstein, Ke Hu, Minho Jin, Anjuli Kannan, Qiao Liang, Ian McGraw, Cal Peyser, Rohit Prabhavalkar, Golan Pundak, David Rybach, Yuan Shangguan, Yash Sheth, Trevor Strohman, Mirko Visontai, Yonghui Wu, Yu Zhang, Ding Zhao

Figure 1 for A Streaming On-Device End-to-End Model Surpassing Server-Side Conventional Model Quality and Latency
Figure 2 for A Streaming On-Device End-to-End Model Surpassing Server-Side Conventional Model Quality and Latency
Figure 3 for A Streaming On-Device End-to-End Model Surpassing Server-Side Conventional Model Quality and Latency
Figure 4 for A Streaming On-Device End-to-End Model Surpassing Server-Side Conventional Model Quality and Latency
* In Proceedings of IEEE ICASSP 2020 
  Access Paper or Ask Questions

Deliberation Model Based Two-Pass End-to-End Speech Recognition


Mar 17, 2020
Ke Hu, Tara N. Sainath, Ruoming Pang, Rohit Prabhavalkar

Figure 1 for Deliberation Model Based Two-Pass End-to-End Speech Recognition
Figure 2 for Deliberation Model Based Two-Pass End-to-End Speech Recognition
Figure 3 for Deliberation Model Based Two-Pass End-to-End Speech Recognition
Figure 4 for Deliberation Model Based Two-Pass End-to-End Speech Recognition

  Access Paper or Ask Questions

Recognizing long-form speech using streaming end-to-end models


Oct 24, 2019
Arun Narayanan, Rohit Prabhavalkar, Chung-Cheng Chiu, David Rybach, Tara N. Sainath, Trevor Strohman

Figure 1 for Recognizing long-form speech using streaming end-to-end models
Figure 2 for Recognizing long-form speech using streaming end-to-end models
Figure 3 for Recognizing long-form speech using streaming end-to-end models
Figure 4 for Recognizing long-form speech using streaming end-to-end models

  Access Paper or Ask Questions

Large-Scale Multilingual Speech Recognition with a Streaming End-to-End Model


Sep 11, 2019
Anjuli Kannan, Arindrima Datta, Tara N. Sainath, Eugene Weinstein, Bhuvana Ramabhadran, Yonghui Wu, Ankur Bapna, Zhifeng Chen, Seungji Lee

Figure 1 for Large-Scale Multilingual Speech Recognition with a Streaming End-to-End Model
Figure 2 for Large-Scale Multilingual Speech Recognition with a Streaming End-to-End Model
Figure 3 for Large-Scale Multilingual Speech Recognition with a Streaming End-to-End Model
* Accepted in Interspeech 2019 
  Access Paper or Ask Questions

Two-Pass End-to-End Speech Recognition


Aug 29, 2019
Tara N. Sainath, Ruoming Pang, David Rybach, Yanzhang He, Rohit Prabhavalkar, Wei Li, Mirkó Visontai, Qiao Liang, Trevor Strohman, Yonghui Wu, Ian McGraw, Chung-Cheng Chiu

Figure 1 for Two-Pass End-to-End Speech Recognition
Figure 2 for Two-Pass End-to-End Speech Recognition
Figure 3 for Two-Pass End-to-End Speech Recognition
Figure 4 for Two-Pass End-to-End Speech Recognition

  Access Paper or Ask Questions

Phoneme-Based Contextualization for Cross-Lingual Speech Recognition in End-to-End Models


Jul 22, 2019
Ke Hu, Antoine Bruguier, Tara N. Sainath, Rohit Prabhavalkar, Golan Pundak

Figure 1 for Phoneme-Based Contextualization for Cross-Lingual Speech Recognition in End-to-End Models
Figure 2 for Phoneme-Based Contextualization for Cross-Lingual Speech Recognition in End-to-End Models
Figure 3 for Phoneme-Based Contextualization for Cross-Lingual Speech Recognition in End-to-End Models
Figure 4 for Phoneme-Based Contextualization for Cross-Lingual Speech Recognition in End-to-End Models

  Access Paper or Ask Questions

Improving Performance of End-to-End ASR on Numeric Sequences


Jul 01, 2019
Cal Peyser, Hao Zhang, Tara N. Sainath, Zelin Wu

Figure 1 for Improving Performance of End-to-End ASR on Numeric Sequences
Figure 2 for Improving Performance of End-to-End ASR on Numeric Sequences
Figure 3 for Improving Performance of End-to-End ASR on Numeric Sequences

  Access Paper or Ask Questions

A spelling correction model for end-to-end speech recognition


Feb 19, 2019
Jinxi Guo, Tara N. Sainath, Ron J. Weiss

Figure 1 for A spelling correction model for end-to-end speech recognition
Figure 2 for A spelling correction model for end-to-end speech recognition
Figure 3 for A spelling correction model for end-to-end speech recognition
Figure 4 for A spelling correction model for end-to-end speech recognition
* Accepted to ICASSP 2019 
  Access Paper or Ask Questions

Streaming End-to-end Speech Recognition For Mobile Devices


Nov 15, 2018
Yanzhang He, Tara N. Sainath, Rohit Prabhavalkar, Ian McGraw, Raziel Alvarez, Ding Zhao, David Rybach, Anjuli Kannan, Yonghui Wu, Ruoming Pang, Qiao Liang, Deepti Bhatia, Yuan Shangguan, Bo Li, Golan Pundak, Khe Chai Sim, Tom Bagby, Shuo-yiin Chang, Kanishka Rao, Alexander Gruenstein

Figure 1 for Streaming End-to-end Speech Recognition For Mobile Devices
Figure 2 for Streaming End-to-end Speech Recognition For Mobile Devices
Figure 3 for Streaming End-to-end Speech Recognition For Mobile Devices
Figure 4 for Streaming End-to-end Speech Recognition For Mobile Devices

  Access Paper or Ask Questions

Contextual Speech Recognition with Difficult Negative Training Examples


Oct 29, 2018
Uri Alon, Golan Pundak, Tara N. Sainath

Figure 1 for Contextual Speech Recognition with Difficult Negative Training Examples
Figure 2 for Contextual Speech Recognition with Difficult Negative Training Examples
Figure 3 for Contextual Speech Recognition with Difficult Negative Training Examples
Figure 4 for Contextual Speech Recognition with Difficult Negative Training Examples

  Access Paper or Ask Questions

Deep context: end-to-end contextual speech recognition


Aug 07, 2018
Golan Pundak, Tara N. Sainath, Rohit Prabhavalkar, Anjuli Kannan, Ding Zhao

Figure 1 for Deep context: end-to-end contextual speech recognition
Figure 2 for Deep context: end-to-end contextual speech recognition
Figure 3 for Deep context: end-to-end contextual speech recognition
Figure 4 for Deep context: end-to-end contextual speech recognition

  Access Paper or Ask Questions

State-of-the-art Speech Recognition With Sequence-to-Sequence Models


Feb 23, 2018
Chung-Cheng Chiu, Tara N. Sainath, Yonghui Wu, Rohit Prabhavalkar, Patrick Nguyen, Zhifeng Chen, Anjuli Kannan, Ron J. Weiss, Kanishka Rao, Ekaterina Gonina, Navdeep Jaitly, Bo Li, Jan Chorowski, Michiel Bacchiani

Figure 1 for State-of-the-art Speech Recognition With Sequence-to-Sequence Models
Figure 2 for State-of-the-art Speech Recognition With Sequence-to-Sequence Models
Figure 3 for State-of-the-art Speech Recognition With Sequence-to-Sequence Models
Figure 4 for State-of-the-art Speech Recognition With Sequence-to-Sequence Models
* ICASSP camera-ready version 
  Access Paper or Ask Questions

Multilingual Speech Recognition With A Single End-To-End Model


Feb 15, 2018
Shubham Toshniwal, Tara N. Sainath, Ron J. Weiss, Bo Li, Pedro Moreno, Eugene Weinstein, Kanishka Rao

Figure 1 for Multilingual Speech Recognition With A Single End-To-End Model
Figure 2 for Multilingual Speech Recognition With A Single End-To-End Model
Figure 3 for Multilingual Speech Recognition With A Single End-To-End Model
Figure 4 for Multilingual Speech Recognition With A Single End-To-End Model
* Accepted in ICASSP 2018 
  Access Paper or Ask Questions

An analysis of incorporating an external language model into a sequence-to-sequence model


Dec 06, 2017
Anjuli Kannan, Yonghui Wu, Patrick Nguyen, Tara N. Sainath, Zhifeng Chen, Rohit Prabhavalkar

Figure 1 for An analysis of incorporating an external language model into a sequence-to-sequence model
Figure 2 for An analysis of incorporating an external language model into a sequence-to-sequence model
Figure 3 for An analysis of incorporating an external language model into a sequence-to-sequence model
Figure 4 for An analysis of incorporating an external language model into a sequence-to-sequence model

  Access Paper or Ask Questions

No Need for a Lexicon? Evaluating the Value of the Pronunciation Lexica in End-to-End Models


Dec 05, 2017
Tara N. Sainath, Rohit Prabhavalkar, Shankar Kumar, Seungji Lee, Anjuli Kannan, David Rybach, Vlad Schogol, Patrick Nguyen, Bo Li, Yonghui Wu, Zhifeng Chen, Chung-Cheng Chiu

Figure 1 for No Need for a Lexicon? Evaluating the Value of the Pronunciation Lexica in End-to-End Models
Figure 2 for No Need for a Lexicon? Evaluating the Value of the Pronunciation Lexica in End-to-End Models
Figure 3 for No Need for a Lexicon? Evaluating the Value of the Pronunciation Lexica in End-to-End Models
Figure 4 for No Need for a Lexicon? Evaluating the Value of the Pronunciation Lexica in End-to-End Models

  Access Paper or Ask Questions

Minimum Word Error Rate Training for Attention-based Sequence-to-Sequence Models


Dec 05, 2017
Rohit Prabhavalkar, Tara N. Sainath, Yonghui Wu, Patrick Nguyen, Zhifeng Chen, Chung-Cheng Chiu, Anjuli Kannan

Figure 1 for Minimum Word Error Rate Training for Attention-based Sequence-to-Sequence Models
Figure 2 for Minimum Word Error Rate Training for Attention-based Sequence-to-Sequence Models
Figure 3 for Minimum Word Error Rate Training for Attention-based Sequence-to-Sequence Models
Figure 4 for Minimum Word Error Rate Training for Attention-based Sequence-to-Sequence Models

  Access Paper or Ask Questions

Improving the Performance of Online Neural Transducer Models


Dec 05, 2017
Tara N. Sainath, Chung-Cheng Chiu, Rohit Prabhavalkar, Anjuli Kannan, Yonghui Wu, Patrick Nguyen, Zhifeng Chen

Figure 1 for Improving the Performance of Online Neural Transducer Models
Figure 2 for Improving the Performance of Online Neural Transducer Models
Figure 3 for Improving the Performance of Online Neural Transducer Models
Figure 4 for Improving the Performance of Online Neural Transducer Models

  Access Paper or Ask Questions