[1]成德源.无结构码的快速最邻近模式搜索[J].深圳大学学报理工版,1990,(1-2):21-26.
 Cheng Deyuan.Fast Nearest Neighbor Pattern Search for Nonstructured Codes[J].Journal of Shenzhen University Science and Engineering,1990,(1-2):21-26.
点击复制

无结构码的快速最邻近模式搜索()
分享到:

《深圳大学学报理工版》[ISSN:1000-2618/CN:44-1401/N]

卷:
期数:
1990年1-2期
页码:
21-26
栏目:
环境与能源
出版日期:
1990-06-30

文章信息/Info

Title:
Fast Nearest Neighbor Pattern Search for Nonstructured Codes
作者:
成德源
深圳大学电子工程系
Author(s):
Cheng Deyuan
Department of electronic engineering, Shenzhen University
摘要:
本文提出二进制超平面试验(BHT)方法,使最邻近模式搜索复杂性随比特率线性增长,远低于传统的随比特率及空间维数均指数增长的EFS方法的搜索复杂性。BHT方法已成功地应用于实时语音波形矢量量化编码。
Abstract:
The demanding exhaustive codebook search can be avoided by more efficient algorithm, the Binary Hyperplane Testing (BHT)algorithm, for fast codebook searching when least Euclidean distance is the appropriate distortion or dissimilarity measure. Let N denote the size of a codebook, and k denote the dimension of a vector. The seartch complexity of the BHT algorithm increases linearly with the bit rate r= ( log2N ) /k. The search complexity o? the BHT algorithm is sufficiently low that it has been used to implement a real-time speech, waveform vector quantizer with sampling rate 8 KHz, dimension k=4,and codebook size N= .1.28 using a single programmable signal processor, the TMS 32010.
更新日期/Last Update: 2016-05-10