2 citations
1 paper
Romain Hennequin, Jimena Royo-Letelier, Manuel Moussallam
In this paper, we propose to infer music genre embeddings from audio datasets carrying semantic information about genres. We show that such embeddings can be used for disambiguatin…