Automatic artist recognition of songs using a hierarchic structure
Songhao Zhu, Yuncai Liu · China-Ireland International Conference on Information and Communications Technologies (CIICT 2007) · 2007
In this paper, we develop a hierarchic structure to complete the reorganization of artist automatically. This proposed system can efficiently identify the artist's voice of a raw song by analyzing substantive features extracted from both pure music and singing song mixed with accompanying music. The silent segments of a raw song are firstly detected and removed. Then, the pure music segments are distinguished from the segments of singing song mixed with accompanying music by support vector machines based on learning from the training data. Furthermore, the singing voice is recognized by the clustering results and the domain knowledge related to properties of the accompanying music and voice. The experiments on different genres of songs illustrate our proposed system is possible.