Audio Artist Identification by Deep Neural Network

Audio Artist Identification (AAI) is an important problem in Music Information Retrieval (MIR). The famous annual competition, Music Information Retrieval Evaluation eXchange (MIREX), also takes AAI task as one of the four training&testing tasks. We built a hybrid model based on Deep Belief Network (DBN) and Stacked Denoising Autoencoder (SDA) to identify the artist from audio signal. As a matter of copyright, sponsors of MIREX can not publish their data set. We built a comparable data set to test our model. We got an accuracy of 76.26% in our data set while the best result reported in MIREX 2012 is 69.70%. We think our method is promising even though we test it in a different data set, since our data set is comparable to that in MIREX by size. We also found some interesting phenomena in our model, like "the Three Bs" gathered together in our model and samples from different classes become farther away from each other when transformed by more layers in our model. We want to report our work in this paper for farther discussion.