关键词: 随机局部搜索 LBG 聚类 矢量量化

Abstract:

The LBG algorithm is one of the common and important methods used in speaker recognition. But the mare drawback of the LBG algorithm is that it often   gets trapped in local optima that are significantly worse than the global optimum. This paper studies a randomized local search algorithm for vector qua ntization. The results indicate that the proposed algorithrn is easy to implement and is competitive compared with the current best clustering methods . In addition, it is demonstrated to be more effective in the clustering for speech parameters, and can obtain better codebook quality compared with the  LBG algorithm. The proposed algorithm in this paper also shows a new idea in designing the best codebook for solvin more complex problems in speaker rec   ognition.

Key words: randomized local search LBG clustering vector quantization