作者: Mohammed El-Henawy , Nemat S. Abdel Kader , Mohammed Abu El-Yazeed
DOI:
关键词: Speech recognition 、 Pattern recognition 、 Computer science 、 Linde–Buzo–Gray algorithm 、 Algorithm 、 Vector quantization 、 Artificial intelligence 、 Distortion 、 Codebook 、 Matching (graph theory) 、 Speedup 、 Identification (information) 、 Learning vector quantization
摘要: This paper introduces an algorithm for speaker identification based on multi-codebook vector quantization (MCVQ). MCVQ combines different size codebooks to achieve high recognition accuracy text-independent and reduce the number of distortion calculations during matching between test frame speakers’ codebooks. Experimental work has shown that proposed model speed up process without approximately decreasing accuracy.