Automatic Music Genre Classification Based on Sparse Representation and Wavelet Packet Transform with Discrete Trigonometric Transform
Shih-Hao Chen, Sung-Yuan Ko, Shi-Huang Chen · 2016
In this paper, an effective music genre classification algorithm using sparse representation based classification (SRC) and wavelet packet transform (WPT) with discrete trigonometric transform (DTT) is developed for improving the classification performance. The first step of the proposed algorithm is to apply moving average filter and Butterworth low-pass filter to partly eliminate the effect of fluctuation in short-term signal. Then one can make use of SRC and WPT with DTT to accurately classify and increase classification performance. Sparse representation based classification has been widely used for music genre classification via the primal-dual algorithm for linear programming to search the most compact representation of the signal in the digital domain. To investigate its performance, the proposed method is validated by comparison with various discrete cosine transform types and classification methods. Various experimental results carried out one the ISMIR 2004 Genre dataset show that the proposed method can achieve higher classification accuracy than other music genre classification methods with the same experimental setup.