Codewords-Expanded Enhanced Residual Vector Quantization for Approximate Nearest Neighbor Search

被引：0
|
作者：
Ai L. [1 ,2 ]
Cheng H. [1 ]
Tao Y. [1 ]
Yu J. [3 ]
Zheng X. [1 ,2 ]
Liu D. [1 ,2 ]
机构：
[1] Department of Computer and Information, Anqing Normal University, Anqing
[2] The University Key Laboratory of Intelligent Perception and Computing of Anhui Province, Anqing
[3] School of Computer Science and Technology, Huazhong University of Science and Technology, Wuhan
来源：
Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao/Journal of Computer-Aided Design and Computer Graphics | 2022年 / 34卷 / 03期
关键词：
Approximate nearest neighbor search; Mean-equisection vector; Residual vector; Vector quantization;
D O I：
10.3724/SP.J.1089.2022.18931
中图分类号：
学科分类号：
摘要：
A codewords-expanded enhanced residual vector quantization (CERVQ) is proposed to improve the accuracy of approximate nearest neighbor (ANN) search for feature vectors. It combines enhanced residual vector quantization (ERVQ) with the method of calculating mean-equisection vector to reduce training error and improves the quantization accuracy. Firstly, the mean-equisection vector is used to compute residual vector as the input to next layer in the training stage, except for the first layer. Based on this, an iterative method for optimizing codebooks is designed to reduce overall quantization error. Secondly, the codebook of each layer is expanded with the mean-equisection vectors to generate new codewords in quantization stage, which are used to quantize input feature vectors to improve quantization accuracy. Finally, a method of calculating asymmetric Euclidean distance is implemented for ANN search. The CERVQ is compared with 5 typical methods on two public SIFT and GIST datasets. Experiment results show that the training error is reducesed by 10%~24% and the recall rate of ANN search is increased by 1%~44%. In addition, the scale of codebook in the CERVQ canbe reduced by 50% under the condition of reaching the same recall rate. © 2022, Beijing China Science Journal Publishing Co. Ltd. All right reserved.
引用
收藏
页码：459 / 469
页数：10
相关论文
共 20 条
  • [1] Pan Z B, Wang L Z, Wang Y, Et al., Product quantization with dual codebooks for approximate nearest neighbor search, Neurocomputing, 401, pp. 59-68, (2020)
  • [2] Li W, Zhang Y, Sun Y F, Et al., Approximate nearest neighbor search on high dimensional data: Experiments, analyses, and improvement, IEEE Transactions on Knowledge and Data Engineering, 32, 8, pp. 1475-1488, (2020)
  • [3] Luan Tingting, Zhu Jihua, Xu Siyu, Et al., Hashing method for image retrieval based on product quantization with Huffman coding, Journal of Image and Graphics, 24, 3, pp. 389-399, (2019)
  • [4] Jegou H, Douze M, Schmid C., Product quantization for nearest neighbor search, IEEE Transactions on Pattern Analysis and Machine Intelligence, 33, 1, pp. 117-128, (2011)
  • [5] Gong Y C, Lazebnik S, Gordo A, Et al., Iterative quantization: a procrustean approach to learning binary codes for large-scale image retrieval, IEEE Transactions on Pattern Analysis and Machine Intelligence, 35, 12, pp. 2916-2929, (2013)
  • [6] Ge T Z, He K M, Ke Q F, Et al., Optimized product quantization, IEEE Transactions on Pattern Analysis and Machine Intelligence, 36, 4, pp. 744-755, (2014)
  • [7] Kalantidis Y, Avrithis Y., Locally optimized product quantization for approximate nearest neighbor search, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2329-2336, (2014)
  • [8] An S, Huang Z B, Bai S, Et al., Quarter-point product quantization for approximate nearest neighbor search, Pattern Recognition Letters, 125, pp. 187-194, (2019)
  • [9] Yang J C, Chen B, Xia S T., Mean-removed product quantization for large-scale image retrieval, Neurocomputing, 406, pp. 77-88, (2020)
  • [10] Babenko A, Lempitsky V., Tree quantization for large-scale similarity search and classification, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 4240-4248, (2015)
12