The VoiceMOS Challenge 2022

The VoiceMOS Challenge 2022 has ended! Material from the challenge will remain available online. Read our summary paper of the challenge here.

Human listening tests are the gold standard for evaluating synthesized speech. Objective measures of speech quality have low correlation with human ratings, and the generalization abilities of current data-driven quality prediction systems suffer significantly from domain mismatch. The VoiceMOS Challenge aims to encourage research in the area of automatic prediction of Mean Opinion Scores (MOS) for synthesized speech. This challenge has two tracks:

Participation is open to all. The main track is required for all participants, and the out-of-domain track is optional. This challenge has preliminarily been accepted as a special session for Interspeech 2022, and participants are strongly encouraged to submit papers to the special session. The focus of the special session is on understanding and comparing MOS prediction techniques using a standardized dataset.


The challenge has ended, but you can still access the CodaLab challenge page here.

You can also find the BVCC dataset that was used in the challenge here.


The schedule for the challenge is as follows: