CHiME-6 Challenge: Tackling Multispeaker Speech Recognition for Unsegmented Recordings

By Shinji Watanabe, Michael I. Mandel, Jon Barker, Emmanuel Vincent. 2020

407 citations recorded by the source corpus.

Following the success of the 1st, 2nd, 3rd, 4th and 5th CHiME challenges we organize the 6th CHiME Speech Separation and Recognition Challenge (CHiME-6). The new challenge revisits the previous CHiME-5 challenge and further considers the problem of distant multi-microphone conversational speech diarization and recognition in everyday home environments. Speech material is the same as the previous CHiME-5 recordings except for accurate array synchronization. The material was elicited using a dinner party scenario with efforts taken to capture data that is representative of natural conversational speech. This paper provides a baseline description of the CHiME-6 challenge for both segmented multispeaker speech recognition (Track 1) and unsegmented multispeaker speech recognition (Track 2). Of note, Track 2 is the first challenge activity in the community to tackle an unsegmented multispeaker speech recognition scenario with a complete set of reproducible open source baselines providing speech enhancement, speaker diarization, and speech recognition modules.

Subfields: Artificial Intelligence, Evaluation & Benchmarks, Speech & Audio, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Evaluation & Benchmarks

In Speech & Audio

In Signal Processing

Citations in the corpus

Robust Speech Recognition via Large-Scale Weak Supervision (2022, methodology): “• CHiME-6 : For CHiME-6 (Watanabe et al., 2020), we downloaded the CHiME-5 dataset and followed the stage 0 of the s5 track1 recipe to create the CHiME-6 dataset which fixes synchronization.”
BigSSL: Exploring the Frontier of Large-Scale Semi-Supervised Learning for Automatic Speech Recognition (2021, methodology): “We then present our results on two public tasks, SpeechStew [30] and CHiME-6 [70] and compare them to the current SoTA benchmarks.”
SpeechStew: Simply Mix All Available Speech Recognition Data to Train One Large Neural Network (2021, positive): “CHiME-6 [7] is a low resource noisy speech dataset.”
AISHELL-4: An Open Source Dataset for Speech Enhancement, Separation, Recognition and Speaker Diarization in Conference Scenario (2021, background): “To help with the data insufficiency problem, several relevant datasets have been collected and released [6, 7, 8, 9, 10].”
Continuous Speech Separation with Conformer (2020, background): “However, when applied to acoustically and linguistically complicated scenarios such as conversation transcription [5, 6], the ASR systems still suffer from the performance limitation due to overlapped speech and quick speaker turn-taking, which break the usually assumed single active speaker conditi”
The STC System for the CHiME-6 Challenge (2020, background): “The CHiME-6 challenge [1] considers the problem of distant multi-microphone conversational speech recognition in everyday home environments.”
ReVISE: Self-Supervised Speech Resynthesis with Visual Input for Universal and Generalized Speech Enhancement (2022, background): “Speech recorded indoor often contains rever-beration, mechanical noise, and overlapping speech from non-target speakers [54].”
Towards a Competitive End-to-End Speech Recognition for CHiME-6 Dinner Party Transcription (2020, methodology): “There is a series of CHiME Challenges, which task is to encourage researchers to solve the problem of speech recognition in such conditions.”
EasyCom: An Augmented Reality Dataset to Support Algorithms for Easy Communication in Noisy Environments (2021, methodology): “The CHiME-5 [8] and CHiME-6 [9] challenge datasets, released in 2018 and re-run in 2019, respectively, contain 32 audio channels of 4 participants having a dinner party in private homes.”
Towards Neural Diarization for Unlimited Numbers of Speakers Using Global and Local Attractors (2021, background): “This assumption does not matter in some ﬁxed-domain scenarios ( e.g ., two-speaker telephone conversations [7], doctor-patient conversations [8], or four-speaker dinner parties [9]), but becomes a crucial problem in the harder scenarios [3].”
Large-Scale Pre-Training of End-to-End Multi-Talker ASR for Meeting Transcription with Single Distant Microphone (2021, methodology): “We calculate the number of mis-recognized words based on the concatenated minimumpermutation word error rate [6].”