Musical genre classification of MPEG-4 TwinVQ audio data

Abstract

We proposed a musical feature based on LSP (Line Spectrum Pair) parameter directly extracted from the bitstream in the MPEG-4 TwinVQ audio data. Our key idea is to extract the musical features by using information stored in the bitstream without decoding to audio signals. In this paper, we propose two musical features for musical genre classification of MPEG-4 TwinVQ audio data. For extracting musical features, we focus on LPC (Linear Predictive Coding) cepstrum and LPC coefficient computed from LSP parameter in the bitstream of TwinVQ audio data by inverse operations of encoding steps. The musical features based on LPC cepstrum and on LPC coefficient are computed by Discrete Wavelet Transform (DWT). For musical genre classification, we use the Discriminant Analysis (DA) as a classifier. We experimented on 2, 196 TwinVQ audio data collected from 10 musical genres and evaluated the performance of the musical features. From the experiments, we got the correct ratios 79.7% and 84.1% for LPC coefficient-based musical feature and LPC cepstrum-based musical feature, respectively. And we compared the performance of two musical features. The experiments showed that LPC cepstrum-based musical feature had good performance for musical genre classification in the compressed domain of MPEG-4 TwinVQ audio compression.