Content-Based Classification of Music Using VQ-Multifeature Clustering Technique

Fuzzy declustering-based vector quantization

“…Many applications of the VQ idea and its extended versions have been found in image analysis and coding [4][5][6][7][8][9][10]; speech compression, speech and speaker recognition [3,[11][12][13]; and other pattern classification and recognition systems [14][15][16][17][18]. In speech recognition, VQ is typically utilized to reduce the information rate of the speech signal to a much lower rate by the use of a codebook with a relatively small number of codewords; VQ codebook can be used as a recognition processor for efficient speech pattern comparison where the main idea is to use linear predictive coding (LPC) to extract the redundancy from speech signals and subsequently use VQ to approximate a residual signal; and as an effective implementation of discrete hidden Markov models in which both time and spectral constraints are used to quantize an entire speech utterance in a well-defined manner [3].…”