Residual Vector Product Quantization for Approximate Nearest Neighbor Search

被引:4
|
作者
Xu, Zhi [1 ]
Niu, Lushuai [1 ]
Meng, Ruimin [1 ]
Zhao, Longyang [1 ]
Ji, Jianqiu [1 ,2 ]
机构
[1] Guilin Univ Elect Technol, Sch Comp Sci & Informat Secur, Guilin 541004, Peoples R China
[2] Doodod Technol Co Ltd, Guilin, Peoples R China
来源
ADVANCES IN KNOWLEDGE DISCOVERY AND DATA MINING, PAKDD 2022, PT I | 2022年 / 13280卷
基金
中国国家自然科学基金;
关键词
Vector quantization; Approximate nearest neighbor search; Residual encoding;
D O I
10.1007/978-3-031-05933-9_17
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Product Quantization is popular for approximate nearest neighbor search, which decomposes the vector space into Cartesian product of several subspaces and constructs separately one codebook for each subspace. The construction of codebooks dominates the quantization error that directly impacts the retrieval accuracy. In this paper, we propose a novel quantization method, residual vector product quantization (RVPQ), which constructs a residual hierarchy structure consisted of several ordered residual codebooks for each subspace. The proposed method minimizes the quantization error by jointly optimizing all the codebooks in each subspace using the efficient mini-batch stochastic gradient descent algorithm. Furthermore, an efficient encoding method, based on H-variable Beam Search, is also proposed to reduce the computation complexity of encoding with negligible loss of accuracy. Extensive experiments show that our proposed method outperforms the-state-of-the-art on retrieval accuracy while retaining a comparable computation complexity.
引用
收藏
页码:208 / 220
页数:13
相关论文
共 50 条
  • [21] Search algorithms for vector quantization and nearest neighbor classification
    Ryan, TW
    Pothier, S
    Pierson, W
    ALGORITHMS FOR SYNTHETIC APERTURE RADAR IMAGERY VIII, 2001, 4382 : 276 - 285
  • [22] Iteratively Multiple Projections Optimization for Product Quantization in Nearest Neighbor Search
    Li, Jin
    Lan, Xuguang
    Zheng, Nanning
    2017 IEEE INTERNATIONAL CONFERENCE ON BIG KNOWLEDGE (IEEE ICBK 2017), 2017, : 65 - 71
  • [23] STACKED PRODUCT QUANTIZATION FOR NEAREST NEIGHBOR SEARCH ON LARGE DATASETS
    Wang, Jun
    Li, Zhiyang
    Du, Yegang
    Qu, Wenyu
    2016 IEEE TRUSTCOM/BIGDATASE/ISPA, 2016, : 1621 - 1627
  • [24] Object detection with a nearest neighbor classifier based on residual vector quantization
    Barnes, CF
    TERRORISM AND COUNTERTERRORISM METHODS AND TECHNOLOGIES, 1997, 2933 : 77 - 85
  • [25] Effective product quantization-based indexing for nearest neighbor search
    Chiu, Chih-Yi
    Chiu, Jih-Sheng
    Markchit, Sarawut
    Chou, Sheng-Hao
    MULTIMEDIA TOOLS AND APPLICATIONS, 2019, 78 (03) : 2877 - 2895
  • [26] Effective product quantization-based indexing for nearest neighbor search
    Chih-Yi Chiu
    Jih-Sheng Chiu
    Sarawut Markchit
    Sheng-Hao Chou
    Multimedia Tools and Applications, 2019, 78 : 2877 - 2895
  • [27] Distance Encoded Product Quantization for Approximate K-Nearest Neighbor Search in High-Dimensional Space
    Heo, Jae-Pil
    Lin, Zhe
    Yoon, Sung-Eui
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2019, 41 (09) : 2084 - 2097
  • [28] Sea mine detection with a nearest neighbor classifier based on residual vector quantization
    Barnes, CF
    DETECTION AND REMEDIATION TECHNOLOGIES FOR MINES AND MINELIKE TARGETS, 1996, 2765 : 167 - 178
  • [29] Optimized high order product quantization for approximate nearest neighbors search
    Linhao Li
    Qinghua Hu
    Frontiers of Computer Science, 2020, 14 : 259 - 272
  • [30] Optimized high order product quantization for approximate nearest neighbors search
    Li, Linhao
    Hu, Qinghua
    FRONTIERS OF COMPUTER SCIENCE, 2020, 14 (02) : 259 - 272