You are not signed in.
My basket (no items)
Sign-in
Sheffield University
  • DESCRIPTION
  • FILES (1)
  • REFERENCES (1)
  • SUPPORT

The CHiME-5 dataset is a collection of over 50 hours of conversational speech recordings collected from twenty real dinner parties that have taken place in real homes. The recordings have been made using multiple 4-channel microphone arrays and have been fully transcribed.

The dataset features:

  • simultaneous recordings from multiple microphone arrays;
  • real conversation, i.e. talkers speaking in a relaxed and unscripted fashion;
  • a range of room acoustics from 20 different homes each with two or three separate recording areas;
  • real domestic noise backgrounds, e.g., kitchen appliances, air conditioning, movement, etc.

Fully-transcribed utterances are provided in continuous audio with ground truth speaker labels and start/end time annotations for segmentation.

The dataset was used for the 5th CHiME Speech Separation and Recognition Challenge. Further information and an open source baseline speech recognition system are available online (http://spandh.dcs.shef.ac.uk/chime_challenge/chime2018).

 

The CHiME-5 speech corpus

The CHiME-5 distant-microphone dinner party speech corpus

OPTIONS
Please check carefully that the terms you select correspond to your intended use of the product.

CHiME-5 data licence - non-commercial 1.00

This is a free of charge academic/non-commercial licence granted for not-for profit organisations.

View Terms

Further details
Term: perpetual (does not require renewal)
Seats: n/a

Price excl. VAT: Free of charge

PLEASE NOTE THAT THE TERMS & CONDITIONS OF THIS LICENCE ARE STANDARD AND THEREFORE NON-NEGOTIABLE

CHiME-5 data licence -commercial 1.00

This is a commercial licence.

View Terms

Further details
Term: perpetual (does not require renewal)
Seats: n/a

Price excl. VAT: £2000.00

Please contact licensing@sheffield.ac.uk if you would like to discuss other options for this licence.