作者: Long-Jhe Yan , Shaw-Hwa Hwang , Shun-Chieh Chang , Chi-Jung Huang
DOI: 10.1109/ISCCSP.2010.5463333
关键词:
摘要: This paper presents an efficient binary search-like algorithm for vector quantization (VQ). The proposed adopts a tree-structured VQ with overlapped codewords (TSOC) to reduce computational complexity and enhance quality. uses expand the scope of search path traverse more appropriate codewords. To further evaluate computations at each stage algorithm, both speech images are considered. With codebook sizes 256, 512 1024, corresponding optimal savings 85.16%, 90.04% 93.46% respectively, compared FSVQ. For speech, reached 51.56% size 128. results indicate that can save significant number computations, depending on codebook.