Semi-supervised Triplet Loss Based Learning of Ambient Audio Embeddings.

Nicolas Turpault, Romain Serizel, Emmanuel Vincent

Browse the full ICASSP paper archive.