Participate in the HEAR 2021 NeurIPS Challenge (Holistic Evaluation of Audio Representations)

[gissel...@yahoo.com]

Dear all,

Participate in the HEAR 2021 NeurIPS Challenge (Holistic Evaluation of Audio Representations)

SUMMARY:

• HEAR 2021 challenges you to develop a general-purpose audio representation that provides a strong basis for learning in a wide variety of tasks and scenarios.
• We will evaluate audio representations using a benchmark suite of secret tasks across a variety of domains, including speech, environmental sound, and music.
• You can submit an existing published model, or new work.
• You have until October 31st 2021 to submit to the challenge.
• PMLR will publish a HEAR special issue. All HEAR participants are invited to submit in January 2022.

Please see our site to learn more about the challenge here: https://neuralaudio.ai/hear2021-holistic-evaluation-of-audio-representations.html

We also welcome any questions you may have!

p.s. Excited by the HEAR competition and believe the audio ML community benefits from broad-scale evaluation? Please forward this email to relevant collaborators and mailing lists! As a new challenge, it helps us a lot if people talk about HEAR.

   The HEAR 2021 Committee

   Bhiksha Raj

   Björn W. Schuller

   Camille Noufi

   Christian James Steinmetz

   Colin Malloy

   Dorien Herremans

   George Tzanetakis

   Gissel Velarde

   Jesse Engel

   Jordie Shier

   Joseph Turian
   Justin Salamon

   Kirk McNally

   Max Henry

   Nicolas Pinto

   Philippe Esling

   Pranay Manocha

   Shinji Watanabe

   Yonatan Bisk

   Zeyu Jin