gmsdi

View the Project on GitHub gladia-research-group/gmsdi

Generalized Multi-Source Inference for Text Conditioned Music Diffusion Models

Accompaniment Generation (MTG-Jamendo Model)

Violin

Conditioning Track Generated Accompaniment Resulting Mix

More samples coming soon…

Accompaniment Generation (Slakh2100 Model)

Given a conditioning track as input, the neural model generates accompanying instruments:

Piano

Conditioning Track Generated Accompaniment Resulting Mix

Guitar

Conditioning Track Generated Accompaniment Resulting Mix

Bass

Conditioning Track Generated Accompaniment Resulting Mix

Drums

Conditioning Track Generated Accompaniment Resulting Mix