Dataset Groups Activity Stream CHiME-4 The CHiME-4 dataset is a large-scale speech recognition dataset, containing over 2 hours of speech from 6 channels. BibTex: @dataset{Emmanuel_Vincent_and_Shinji_Watanabe_and_Aditya_Arie_Nugraha_and_Jon_Barker_and_Ricard_Marxer_2024, abstract = {The CHiME-4 dataset is a large-scale speech recognition dataset, containing over 2 hours of speech from 6 channels.}, author = {Emmanuel Vincent and Shinji Watanabe and Aditya Arie Nugraha and Jon Barker and Ricard Marxer}, doi = {10.57702/3en7tlmr}, institution = {No Organization}, keyword = {'Audio', 'CHiME-4', 'Distant-Talking', 'Noisy Speech', 'Speech Recognition'}, month = {dec}, publisher = {TIB}, title = {CHiME-4}, url = {https://service.tib.eu/ldmservice/dataset/chime-4}, year = {2024} }
