close this message
arXiv smileybones

arXiv Is Hiring a DevOps Engineer

Work on one of the world's most important websites and make an impact on open science.

View Jobs
Skip to main content
Cornell University

arXiv Is Hiring a DevOps Engineer

View Jobs
We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate
arxiv logo > cs > arXiv:1712.01769

Help | Advanced Search

arXiv logo
Cornell University Logo

quick links

  • Login
  • Help Pages
  • About

Computer Science > Computation and Language

arXiv:1712.01769 (cs)
[Submitted on 5 Dec 2017 (v1), last revised 23 Feb 2018 (this version, v6)]

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

Authors: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
View a PDF of the paper titled State-of-the-art Speech Recognition With Sequence-to-Sequence Models, by Chung-Cheng Chiu and 13 other authors
View PDF
Abstract:Attention-based encoder-decoder architectures such as Listen, Attend, and Spell (LAS), subsume the acoustic, pronunciation and language model components of a traditional automatic speech recognition (ASR) system into a single neural network. In previous work, we have shown that such architectures are comparable to state-of-theart ASR systems on dictation tasks, but it was not clear if such architectures would be practical for more challenging tasks such as voice search. In this work, we explore a variety of structural and optimization improvements to our LAS model which significantly improve performance. On the structural side, we show that word piece models can be used instead of graphemes. We also introduce a multi-head attention architecture, which offers improvements over the commonly-used single-head attention. On the optimization side, we explore synchronous training, scheduled sampling, label smoothing, and minimum word error rate optimization, which are all shown to improve accuracy. We present results with a unidirectional LSTM encoder for streaming recognition. On a 12, 500 hour voice search task, we find that the proposed changes improve the WER from 9.2% to 5.6%, while the best conventional system achieves 6.7%; on a dictation task our model achieves a WER of 4.1% compared to 5% for the conventional system.
Comments: ICASSP camera-ready version
Subjects: Computation and Language (cs.CL); Sound (cs.SD); Audio and Speech Processing (eess.AS); Machine Learning (stat.ML)
Cite as: arXiv:1712.01769 [cs.CL]
  (or arXiv:1712.01769v6 [cs.CL] for this version)
  https://doi.org/10.48550/arXiv.1712.01769
arXiv-issued DOI via DataCite

Submission history

From: Chung-Cheng Chiu [view email]
[v1] Tue, 5 Dec 2017 17:24:05 UTC (85 KB)
[v2] Wed, 13 Dec 2017 23:25:23 UTC (85 KB)
[v3] Fri, 15 Dec 2017 18:41:01 UTC (55 KB)
[v4] Fri, 22 Dec 2017 22:55:44 UTC (59 KB)
[v5] Thu, 18 Jan 2018 18:25:33 UTC (59 KB)
[v6] Fri, 23 Feb 2018 18:44:30 UTC (59 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled State-of-the-art Speech Recognition With Sequence-to-Sequence Models, by Chung-Cheng Chiu and 13 other authors
  • View PDF
  • TeX Source
  • Other Formats
view license
Current browse context:
cs.CL
< prev   |   next >
new | recent | 2017-12
Change to browse by:
cs
cs.SD
eess
eess.AS
stat
stat.ML

References & Citations

  • NASA ADS
  • Google Scholar
  • Semantic Scholar

DBLP - CS Bibliography

listing | bibtex
Chung-Cheng Chiu
Tara N. Sainath
Yonghui Wu
Rohit Prabhavalkar
Patrick Nguyen
…
a export BibTeX citation Loading...

BibTeX formatted citation

×
Data provided by:

Bookmark

BibSonomy logo Reddit logo

Bibliographic and Citation Tools

Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)

Code, Data and Media Associated with this Article

alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)

Demos

Replicate (What is Replicate?)
Hugging Face Spaces (What is Spaces?)
TXYZ.AI (What is TXYZ.AI?)

Recommenders and Search Tools

Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
  • Author
  • Venue
  • Institution
  • Topic

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
  • About
  • Help
  • contact arXivClick here to contact arXiv Contact
  • subscribe to arXiv mailingsClick here to subscribe Subscribe
  • Copyright
  • Privacy Policy
  • Web Accessibility Assistance
  • arXiv Operational Status
    Get status notifications via email or slack