This is a reproduction of the paper 'LEARNING AUDIO FEATURES FOR SINGER IDENTIFICATION AND EMBEDDING' submitted to ICLR2018 - View it on GitHub