Music Genre Classification using Support Vector Machine and Empirical Mode Decomposition
Eamin Chaudary, Sumair Aziz, Muhammad Umar Khan, Paul Gretschmann · 2021
Classification of music classes is very tricky in the field of music information retrieval (MIR). In this article, a novel approach is proposed first, the audio signal is denoised and region of interest is extracted using Empirical Mode Decomposition (EMD) and first eight Intrinsic Mode Functions (IMFs) are selected and time and frequency domain features are extracted then linear Support Vector Machine (SVM-L) is trained to get the best accuracy of 94.0% with 5 genre i.e. blues, classic, hip-hop, metal, and pop.