Singer identification using time-frequency audio feature

Authors:
Pafan Doungpaisan
Affiliations:
Department of Information Technology, Faculty of Industrial Technology and Management, King Mongkut's University of Technology North Bangkok, Bangsue, Bangkok, Thailand
Venue:
ISNN'11 Proceedings of the 8th international conference on Advances in neural networks - Volume Part II
Year:
2011

Citing 6
Cited 0

Speech/music segmentation using entropy and dynamism features in a HMM classification framework

Speech Communication
Environmental Sound Recognition by Multilayered Neural Networks

CIT '04 Proceedings of the The Fourth International Conference on Computer and Information Technology
Generic Audio Classification Using a Hybrid Model Based on GMMs and HMMs

MMM '05 Proceedings of the 11th International Multimedia Modelling Conference
Singing voice recognition based on matching of spectrogram pattern

IJCNN'09 Proceedings of the 2009 international joint conference on Neural Networks
A modeling of singing voice robust to accompaniment sounds and its application to singer identification and vocal-timbre-similarity-based music information retrieval

IEEE Transactions on Audio, Speech, and Language Processing
Drum Sound Recognition for Polyphonic Audio Signals by Adaptation and Matching of Spectrogram Templates With Harmonic Structure Suppression

IEEE Transactions on Audio, Speech, and Language Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Singer identification is a difficult topic in music information Retriveal research area. Because the background instrumental accompaniment in audio music is regarded as noise source that has to reduce a performance. This paper proposes a singer identification algorithm thai is able to automatically identify a singer in an audio music signal with background music by using Time-Frequency audio feature. The main idea is used a spectrogram to able effective Time-Frequency feature and used as the input for classification. The proposed technique is test with 20 different singer. Sereval classification technique are compared,such as Feed-Forward Neural Network, k-Nearest Neighbor (kNN) and Minimum least square linear classifier(Fisher). The experimental result on singer identification using a spectrogram with Feed-Forward Neural Networkand and k-Nearest Neighbor (kNN) can effectively identify the singer in music signal with background music more than 92%.