Popular music analysis: Chorus and emotion Detection
Chia-Hung Yeh, Yu Dun Lin, Ming‐Sui Lee, Wen-Yu Tseng · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference · 2010
In this paper, a chorus detection and an e motion detection algorithm for popular music are proposed. First, a popular music is decomposed into chorus and verse segments based on its color representation and MFCCs (Mel-frequency cepstral coefficients ). Four features including intensity, tempo and rhythm regularity are extracted from these structured segments for emotion detection. The emotion of a so ng is classified into four classes of emotions: happy, an gry, depressed and relaxed via a back-propagation neural network classifier. Experimental results show that the average recall an d precision of the proposed chorus detection are approximated to 95% and 84%, respectively; the average precision rate of emotion detection is 88.3% for a test database consisting o f 210 popular music songs. Keyword: Chorus, MFCC, music emotion, neural network I. INTRODUCTION