Publication result detail

Improving Noise Robustness of Automatic Speech Recognition via Parallel Data and Teacher-student Learning

MOŠNER, L.; WU, M.; RAJU, A.; PARTHASARATHI, S.; KUMATANI, K.; SUNDARAM, S.; MAAS, R.; HOFFMEISTER, B.

Original Title

Improving Noise Robustness of Automatic Speech Recognition via Parallel Data and Teacher-student Learning

English Title

Improving Noise Robustness of Automatic Speech Recognition via Parallel Data and Teacher-student Learning

Type

Paper in proceedings (conference paper)

Original Abstract

For real-world speech recognition applications, noise robustnessis still a challenge. In this work, we adopt the teacherstudent(T/S) learning technique using a parallel clean andnoisy corpus for improving automatic speech recognition(ASR) performance under multimedia noise. On top of that,we apply a logits selection method which only preserves the khighest values to prevent wrong emphasis of knowledge fromthe teacher and to reduce bandwidth needed for transferringdata. We incorporate up to 8000 hours of untranscribed datafor training and present our results on sequence trained modelsapart from cross entropy trained ones. The best sequencetrained student model yields relative word error rate (WER)reductions of approximately 10.1%, 28.7% and 19.6% on ourclean, simulated noisy and real test sets respectively comparingto a sequence trained teacher.

English abstract

For real-world speech recognition applications, noise robustnessis still a challenge. In this work, we adopt the teacherstudent(T/S) learning technique using a parallel clean andnoisy corpus for improving automatic speech recognition(ASR) performance under multimedia noise. On top of that,we apply a logits selection method which only preserves the khighest values to prevent wrong emphasis of knowledge fromthe teacher and to reduce bandwidth needed for transferringdata. We incorporate up to 8000 hours of untranscribed datafor training and present our results on sequence trained modelsapart from cross entropy trained ones. The best sequencetrained student model yields relative word error rate (WER)reductions of approximately 10.1%, 28.7% and 19.6% on ourclean, simulated noisy and real test sets respectively comparingto a sequence trained teacher.

Keywords

automatic speech recognition, noise robustness,teacher-student training, domain adaptation

Key words in English

automatic speech recognition, noise robustness,teacher-student training, domain adaptation

Authors

MOŠNER, L.; WU, M.; RAJU, A.; PARTHASARATHI, S.; KUMATANI, K.; SUNDARAM, S.; MAAS, R.; HOFFMEISTER, B.

RIV year

2020

Released

12.05.2019

Publisher

IEEE Signal Processing Society

Location

Brighton

ISBN

978-1-5386-4658-8

Book

Proceedings of ICASSP

Pages from

6475

Pages to

6479

Pages count

5

URL

https://ieeexplore.ieee.org/document/8683422

BibTex

@inproceedings{BUT160006,
  author="MOŠNER, L. and WU, M. and RAJU, A. and PARTHASARATHI, S. and KUMATANI, K. and SUNDARAM, S. and MAAS, R. and HOFFMEISTER, B.",
  title="Improving Noise Robustness of Automatic Speech Recognition via Parallel Data and Teacher-student Learning",
  booktitle="Proceedings of ICASSP",
  year="2019",
  pages="6475--6479",
  publisher="IEEE Signal Processing Society",
  address="Brighton",
  doi="10.1109/ICASSP.2019.8683422",
  isbn="978-1-5386-4658-8",
  url="https://ieeexplore.ieee.org/document/8683422"
}

Documents

mosner_icassp2019_0006475