ModelContrastive Learning of Musical Representations
CLMR
A contrastive learning-based music representation model that takes raw audio as its sole input and learns acoustic feature representations of music through self-supervised contrastive learning. It is frequently used as a baseline method for music information retrieval tasks.
Timeline (last 90 days)
Oct 6
Sona uses the audio-only CLMR as a comparison baseline to examine the gains of audio-plus-metadata-plus-behavioral contrastive learning over purely audio-based representations.
Unverified50%