Announcing the CHiME-5 Challenge

[Shinji Watanabe]

---------------------------------------------
5th CHiME Speech Separation
and Recognition Challenge

Deadline: August 3, 2018

Workshop: Microsoft, Hyderabad, Sep 7, 2018

http://spandh.dcs.shef.ac.uk/chime_challenge/
----------------------------------------------


Dear colleague,

It gives us great pleasure to announce the official launch of the CHiME-5 Challenge.

CHiME-5 considers the problem of distant-microphone conversational speech recognition in everyday home environments. Speech material was elicited using a dinner party scenario with efforts taken to capture data that is representative of natural conversational speech. Participants may use a single microphone array or multiple distributed arrays.


FORUM

If you are considering participating, please join the CHiME-5 Google group for discussions and further announcements.
https://groups.google.com/forum/#!forum/chime5/join


MATERIALS

The Challenge website is now live and contains all the information and data that you will need for participation:
- a detailed description of the challenge scenario and recording conditions,
- real training and development data,
- full instructions for participation and submission.

Baseline software for array synchronization, speech enhancement, and state-of-the-art speech recognition will be provided on March 12.

If you have a question that isn't answered by the website and you expect other participants to have the answer or to be interested by the answer, please post it on the forum. Otherwise, please email us: chimech...@gmail.com.

We look forward to your participation.


IMPORTANT DATES

5th March, 2018 — Training and development set data released
12th March, 2018 — Baseline recognition system released
10th June, 2018 — Workshop registration open
June/July, 2018 — Test data released
3rd Aug, 2018 — Extended abstract and challenge submission deadline
20th Aug, 2018 — Author notification
31st August, 2018 — Workshop registration deadline
7th Sept, 2018 — CHiME-5 Workshop (satellite of Interspeech 2018) and release of results
8th Oct, 2018 — Final paper (2 to 6 pages)


ORGANISERS

Jon Barker, University of Sheffield
Shinji Watanabe, Johns Hopkins University
Emmanuel Vincent, Inria


SPONSORS

Google
Microsoft Research


SUPPORTED BY

International Speech Communication Association (ISCA)
ISCA Robust Speech Processing SIG