Evaluation of Speech Representations for MOS prediction

1 code implementation • 16 Jun 2023 • Frederico S. Oliveira, Edresson Casanova, Arnaldo Cândido Júnior, Lucas R. S. Gris, Anderson S. Soares, Arlindo R. Galvão Filho

Among the supervised and self-supervised learning models using BRSpeechMOS, Whisper-Small achieved the best linear correlation of 0. 6980, and the speaker verification model, SpeakerNet, had linear correlation of 0. 6963.

Self-Supervised Learning Speaker Verification