Neural Audio AI


We believe that machine learning for audio should connect all subcommunities of audio ML researchers (speech, music, environmental sounds, etc.). Moreover, we encourage cross-polination of audio ML research with adjacent fields, such as vision, deep learning, traditional DSP, NLP, and beyond. We specifically encourage benchmarking on audio ML tasks that have high societal impact, in addition to our broader mission of promoting cross-domain evaluation and knowledge sharing.


We will be hosting a NeurIPS challenge entitled “HEAR 2021: Holistic Evaluation of Audio Representations”.

Please consider subscribing to our low-volume announcement mailing list and following us on twitter. Or simply check this web-page for updates.


Do you want to talk about audio ML? Please join our forum. Discussion is open to all audio researchers, whether you are participating in the shared task or simply want to exchange knowledge on audio ML.