The DOI matches its Crossref record; 31 of the 35 references it cites were found in a registry or an index.
Product Quantization for Nearest Neighbor Search
Open accessAbstract
This paper introduces a product quantization-based approach for approximate nearest neighbor search. The idea is to decompose the space into a Cartesian product of low-dimensional subspaces and to quantize each subspace separately. A vector is represented by a short code composed of its subspace quantization indices. The euclidean distance between two vectors can be efficiently estimated from their codes. An asymmetric version increases precision, as it computes the approximate distance between a vector and a code. Experimental results show that our approach searches for nearest neighbors efficiently, in particular in combination with an inverted file system. Results for SIFT and GIST image descriptors show excellent search accuracy, outperforming three state-of-the-art approaches. The scalability of our approach is validated on a data set of two billion vectors.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The text we found for this paper looks like another document, so the self-check wasn’t run.
Citations
Cites
Distinctive Image Features from Scale-Invariant Keypoints
Lowe, D. G. (2004) · International Journal of Computer Vision · doi:10.1023/b:visi.0000029664.99615.94 · Cited by 55,589 · cited 1×
“An exhaustive comparison of the query vector with all codes is prohibitive for very large datasets.”
BackgroundVideo Google: a text retrieval approach to object matching in videos
Sivic, & Zisserman (2003) · Proceedings Ninth IEEE International Conference on Computer Vision · doi:10.1109/iccv.2003.1238663 · Cited by 6,367 · cited 3×
“Similarly to the “Video-Google” approach [24], a codebook is”
BackgroundMethodInfluentialModeling the Shape of the Scene: A Holistic Representation of the Spatial Envelope
Oliva, A., & Torralba, A. B. (2001) · International Journal of Computer Vision · doi:10.1023/a:1011139631724 · Cited by 6,321
Scalable Recognition with a Vocabulary Tree
Nistér, D., & Stewénius, H. (2006) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2006.264 · Cited by 3,641 · cited 2×
“Lattice quantizers offer better quantization properties for uniform vector distributions, but this condition is rarely satisfied by real world vectors.”
BackgroundMethod(1999) · Cited by 3,186
Locality-sensitive hashing scheme based on p-stable distributions
Datar, M., Immorlica, N., Indyk, P., & Mirrokni, V. (2004) · Annual Symposium on Computational Geometry (SoCG) · doi:10.1145/997817.997857 · Cited by 2,971 · cited 9×
“In this case, one of the most popular ANN algorithms is the Euclidean Locality-Sensitive Hashing (E2LSH) [7], [11], which provides theoretical guarantees on the search quality with limited assumptions.”
BackgroundMethodInfluentialObject retrieval with large vocabularies and fast spatial matching
Philbin, J. F., Chum, O., Isard, M., Šivic, J., & Zisserman, A. (2007) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2007.383172 · Cited by 2,931
An Algorithm for Finding Best Matches in Logarithmic Expected Time
Friedman, J. H., Bentley, J. L., & Finkel, R. A. (1977) · ACM Transactions on Mathematical Software · doi:10.1145/355744.355745 · Cited by 2,802 · cited 1×
“( Several multi-dimensional indexing methods, such as the popular KD-tree [5] or other branch and bound techniques, have been proposed to reduce the search time.”
MethodFAST APPROXIMATE NEAREST NEIGHBORS WITH AUTOMATIC ALGORITHM CONFIGURATION
(2009) · Proceedings of the Fourth International Conference on Computer Vision Theory and Applications · doi:10.5220/0001787803310340 · Cited by 2,640 · cited 3×
“An evaluation of the parameters of our approach and a comparison with the state of the art is given in Section V. II.”
BackgroundMethodWhen Is ''Nearest Neighbor'' Meaningful?
Beyer, K., Goldstein, J., Ramakrishnan, R., & Shaft, U. (1999) · Lecture Notes in Computer Science · doi:10.1007/3-540-49257-7_15 · Cited by 2,403
Research works printed but not linked: 2
- Weber, A Quantitative Analysis and Performance Study for Similarity-Search Methods in High-Dimensional Spaces, 194
- Knuth, 1998, The Art of Computer Programming, Sorting and Searching, 3
Cited by
Sorted by citations
Items 901–1,000 of 1,006
Bilinear Optimized Product Quantization for Scalable Visual Content Analysis
Yu, L., Huang, Z. H., Shen, F., Song, J., Shen, H. T., & Zhou, X. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2017.2722224 · Cited by 13
Hinami, R., & Satoh, S. (2018) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/d18-1281 · Cited by 16
“For this, we use the IVFADC system (Jégou et al., 2011) in the manner described in (Hinami and Satoh, 2016).”
MethodFast Approximate Hubness Reduction for Large High-Dimensional Data
Feldbauer, R., Leodolter, M., Plant, C., & Flexer, A. (2018) · IEEE International Conference on Big Knowledge (ICBK) · doi:10.1109/icbk.2018.00055 · Cited by 16
“A plethora of ANN methods has been developed, most of which are based on tree algorithms [21], localitysensitive hashing [22], product quantization [23], or proximity graphs [24].”
MethodCo-weighting semantic convolutional features for object retrieval
Zhu, J., Wang, J., Pang, S., Guan, W., Li, Z., Li, Y., & Qian, X. (2019) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2019.06.006 · Cited by 16
Dual Asymmetric Deep Hashing Learning
Li, J., Zhang, B., Lu, G., & Zhang, D. D. (2019) · IEEE Access · doi:10.1109/access.2019.2927524 · Cited by 18
“Additionally, the method in [21] makes a quantization by decomposing the input space into a Cartesian product of low-dimensional subspaces, dramatically reducing the quantization noise.”
BackgroundMethodSimilarity query processing for high-dimensional data
Qin, J., Wang, W., Xiao, C., et al. (2020) · Proceedings of the VLDB Endowment · doi:10.14778/3415478.3415564 · Cited by 16
“Another line of methods is based on product quantization [29, 19, 32, 24], with the unique ability to handle billions of objects.”
BackgroundRotation-aware representation learning for remote sensing image retrieval
Wu, Z., Zou, C., Wang, Y., Tan, M., & Weise, T. (2021) · Information Sciences · doi:10.1016/j.ins.2021.04.078 · Cited by 16
OSCAR-Net: Object-centric Scene Graph Attention for Image Attribution
Nguyen, E., Bui, T., Swaminathan, V., & Collomosse, J. P. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.01423 · Cited by 18
“For offline hashing, we train a OSCAR-Net model outputting 256-D continuous features using the content loss LC only, then apply PQ [27] for a desirable code length.”
BackgroundMethodMa, X., Li, M., Sun, K., Xin, J., & Lin, J. (2021) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2021.emnlp-main.227 · Cited by 24
“Each sub-vector is quantized using k-means and eventually stored with t bits (Jégou et al., 2011).”
MethodLarge-scale parallel similarity search with Product Quantization for online multimedia services
Andrade, G., Fernandes, A., Gomes, J. M., et al. (2018) · Journal of Parallel and Distributed Computing · doi:10.1016/j.jpdc.2018.11.009 · Cited by 15
Binary multi-view perceptual hashing for image authentication
Du, L., Chen, Z., & Ho, A. T. S. (2020) · Multimedia Tools and Applications · doi:10.1007/s11042-020-08736-6 · Cited by 16
“Recent research has shown that data dependent methods with learning can lead to high quality hashing, such as Spectral Hashing [64], Kernelized LSH (KLSH) [17], Semantic Hashing [42], Product Quantization [11], and Anchor graph hashing [20], LDAHash [44], Semi-supervised Hashing (SSH) [60], Weakly-Supervised Hashing […”
BackgroundInner Product Similarity Search using Compositional CodesPreprint
Du, C., & Wang, J. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1406.4966 · Cited by 16
“We conduct the inner product similarity search experiments over four data sets: SIFT1 M [6], SIFT1 B [25], linear models (LM1 M ), and Netflix.”
BackgroundMethodInfluentialTowards Neural Theorem Proving at ScalePreprint
Minervini, P., Bošnjak, M., Rocktäschel, T., & Riedel, S. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1807.08204 · Cited by 22
“Several families of ANNS algorithms exist, such as Locality-Sensitive Hashing (LSH) (Andoni et al., 2015), Product Quantization (PQ) (Jégou et al., 2011), and Proximity Graphs (PGs) (Malkov et al., 2014).”
BackgroundMethodMIC: Mining Interclass Characteristics for Improved Metric LearningPreprint
Roth, K., Brattoli, B., & Ommer, B. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.11574 · Cited by 45
“We report the Recall@k metric [16] to evaluate image retrieval and the normalized mutual information score (NMI) [22] for the clustering quality.”
MethodRecent Advances in Discrete Speech Tokens: A Review
Guo, Y., Li, Z., Wang, H., Li, B., Shao, C., Zhang, H., Du, C., Xie, C., Liu, S., & Yu, K. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2025.3643619 · Cited by 58
“As quantization is simply done via rounding to integers, there are no explicit codebooks associated with the FSQ process.”
Enhanced Retrieval and Browsing in the IMOTION System
Rossetto, L., Giangreco, I., Tănase, C., et al. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-51814-5_43 · Cited by 15
ACAE-REMIND for online continual learning with compressed feature replay
Wang, K., van de Weijer, J., & Herranz, L. (2021) · Pattern Recognition Letters · doi:10.1016/j.patrec.2021.06.025 · Cited by 15
“Product quantization [13] is an effective quantization method that performs a decomposition of a high-dimensional space into the Cartesian product of a series of subspaces, and quantizes them separately.”
BackgroundMethodInfluentialYu, Z., Li, S., Sun, L., Liu, L., & Haining, W. (2022) · Connection Science · doi:10.1080/09540091.2021.2024510 · Cited by 15
“Themethod focuses on the Euclidean distance between the activation values before and after the quantisation instead of the Euclidean distance between the weights before and after the quantisation in tradition PQ (Jégou et al., 2011), Codebook C = c[1], .”
BackgroundZhu, X., Wang, J., Fang, Z., Yin, X., & Liu, S. (2022) · Applied Sciences · doi:10.3390/app13010413 · Cited by 15
A Dense Representation Framework for Lexical and Semantic Matching
Lin, S.-C., & Lin, J. (2023) · ACM Transactions on Information Systems · doi:10.1145/3582426 · Cited by 23
“PQ [25], OPQ [15], LSH [21] ✓ ✗ JPQ [53], RepCONC [55] ✓ ✓ SPAR [4] ✗ ✓ Our work ✗ ✗”
MethodOrthonormal product quantization network for scalable face image retrieval
Zhang, M., Zhe, X., & Yan, H. (2023) · Pattern Recognition · doi:10.1016/j.patcog.2023.109671 · Cited by 15
ARKGraph: All-Range Approximate K-Nearest-Neighbor Graph
Zuo, C., & Deng, D. (2023) · Proceedings of the VLDB Endowment · doi:10.14778/3603581.3603601 · Cited by 20
“Locality Sensitive Hashing (LSH) [4, 5, 14, 20, 22, 28, 29, 38, 49, 52, 53], product quantization (PQ) [3, 6, 21, 31, 37, 42], and proximity graphs [19, 40] are the cornerstone of almost all the existing index structures and algorithms for ANNS.”
BackgroundPyramid hybrid pooling quantization for efficient fine-grained image retrieval
Zeng, Z., Wang, J., Chen, B., Dai, T., Xia, S., & Wang, Z. (2024) · Pattern Recognition Letters · doi:10.1016/j.patrec.2023.12.022 · Cited by 18
“Compared Baselines : ( i ) 12 coarse-grained hashing methods based on ResNet-18, LSH [29], SH [30], PQ [4], ITQ [31], OPQ [5], DPSH [32], DTH [33], DSH [6], HashNet [28], ADSH [34], PQN [7], and CSQ [17].”
BackgroundMethodInfluentialWang, J., Zeng, Z., Chen, B., Wang, Y., Liao, D., Li, G., Wang, Y., & Xia, S. (2024) · International Journal of Computer Vision · doi:10.1007/s11263-024-02009-7 · Cited by 18
“Quantization (Jégou et al., 2011) divides embedding space into disjoint clusters and approximates each data point by the nearest centroid.”
Efficient Multi-vector Dense Retrieval with Bit Vectors
Nardini, F. M., Rulli, C., & Venturini, R. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-56060-6_1 · Cited by 18
TigerVector: Supporting Vector Search in Graph Databases for Advanced RAGs
Liu, S., Zeng, Z., Chen, L., et al. (2025) · Proceedings - ACM-SIGMOD International Conference on Management of Data · doi:10.1145/3722212.3724456 · Cited by 15
“2, "Carl" , "Anyone in NY?" process periodically cleans up the accumulated deltas and builds a snapshot up to a particular TID.”
Compact Global Descriptors for Visual Search
Chandrasekhar, V., Lin, J., Morère, O., Veillard, A., & Goh, H. (2015) · DCC · doi:10.1109/dcc.2015.54 · Cited by 15
“We provide a thorough evaluation of several state-of-the-art compression schemes based on PCA, Locality Sensitive Hashing, Product Quantization and greedy bit selection, and show that the proposed compression scheme outperforms all existing schemes.”
MethodLinear distance metric learning for large-scale generic image recognition
中山, 英. (2011) · Medical Entomology and Zoology · Cited by 15
Scalability and Total Recall with Fast CoveringLSH
Pham, N., & Pagh, R. (2016) · doi:10.1145/2983323.2983742 · Cited by 15
“In content-based image retrieval systems, a standard approach is to learn short binary codes to represent image objects such that the Hamming distance between codes reflects their neighborhood or semantic similarity in the original space [16, 30, 36, 38].”
BackgroundFast MPEG-CDVS Encoder With GPU-CPU Hybrid Computing
Duan, L., Sun, W., Zhang, X., Wang, S., Chen, J., Yin, J., See, S., Huang, T., Kot, A. C., & Gao, W. (2018) · IEEE Transactions on Image Processing · doi:10.1109/tip.2018.2794203 · Cited by 15
“In early stage of CDVS, the Test Model under Consideration (TMuC) for CDVS [33] employed tree structured vector quantization (TSVQ) [34] and product quantization (PQ) [35] to make compact descriptors.”
Principal Component Analysis Based Filtering for Scalable, High Precision k-NN Search
Feng, H., Eyers, D. M., Mills, S. J., Wu, Y., & Huang, Z. (2017) · IEEE Transactions on Computers · doi:10.1109/tc.2017.2748131 · Cited by 15
“We separately analyse a dataset called “GIST1M” [12] in Section 5.”
BackgroundMethodInfluentialDisentangling Structure and Aesthetics for Style-Aware Image Completion
Gilbert, A., Collomosse, J. P., Jin, H., et al. (2018) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2018.00198 · Cited by 15
“) indicates a product quantization (PQ) [11] of the 256-D search vector to a compact binary form (64 bits) using basis B learned over 0.”
MethodFast Similarity Search via Optimal Sparse Lifting
Li, W., Mao, J., Zhang, Y., & Cui, S. R. (2018) · neural information processing systems · Cited by 16
Semi-supervised Deep Quantization for Cross-modal Search
Wang, X., Zhu, W., & Liu, C. (2019) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3343031.3350934 · Cited by 15
“Quantization [8, 10], as a fairly new compact coding solution, has been developed for cross-modal similarity search in recent years [41].”
BackgroundIsosceles Constraints for Person Re-Identification
Xu, F., Ma, B., Chang, H., & Shan, S. (2020) · IEEE Transactions on Image Processing · doi:10.1109/tip.2020.3020648 · Cited by 15
“We use two standard metrics, the normalized mutual information (NMI) [80] and Recall@K [81], to measure the performances.”
MethodUltra-High Dimensional Sparse Representations with Binarization for Efficient Text Retrieval
Jang, K.-R., Kang, J., Hong, G., Myaeng, S.-H., Park, J., Yoon, T., & Seo, H.-C. (2021) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2021.emnlp-main.78 · Cited by 17
“Our goal is to build an efficient neural ranker without relying on external efficiency-enhancing measures like Approximate Nearest Neighbor (ANN) (Johnson et al., 2019; Jegou et al., 2010).”
MethodDoshi, I., Das, D., Bhutani, A., Kumar, R., Bhatt, R., & Balasubramanian, N. (2021) · Proceedings of the VLDB Endowment · doi:10.14778/3503585.3503594 · Cited by 14
“igh dimensional (e.g., 128, 256, or 2048 dimensional) data sets. Broadly, nearest neighbor search approaches fall into four categories. They can be tree based[5, 8, 19, 23], product quantization based[10, 11, 15, 16], Locality Sensitive Hashing (LSH) based[1, 2, 4, 12, 26], or graph based[9, 13, 17, 18]. Most of the h…”
BackgroundMethodNested Sparse Quantization for Efficient Feature Coding
Boix, X., Roig, G., Leistner, C., & Gool, L. V. (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-33709-3_53 · Cited by 15
“Hashing-based methods tackle fast distance computation rather than feature encoding, and have been typically used to speed up image retrieval in large-scale databases [24,25,26,27].”
MethodMultimedia for Cultural Heritage: Key Issues
Cucchiara, R., Grana, C., Borghesani, D., et al. (2012) · Communications in computer and information science · doi:10.1007/978-3-642-27978-2_18 · Cited by 15
“Modern technology, in terms of multimedia content analysis [14–17], large scale retrieval [18, 19], multimodal aggregation of information sources and annotations [20] and finally database access and management approaches are available.”
BackgroundBoosting k-NN for Categorization of Natural Scenes
Nock, R., Piro, P., Nielsen, F., et al. (2012) · International Journal of Computer Vision · doi:10.1007/s11263-012-0539-2 · Cited by 15
Deep linear discriminant analysis hashing for image retrieval
Yan, L., Lu, H., Wang, C., et al. (2018) · Multimedia Tools and Applications · doi:10.1007/s11042-018-6855-y · Cited by 15
“image classification [7, 9, 30], object detection [29], face recognition [28], and many other vision tasks [25].”
Chen, W., Chen, J., Zou, F., Li, Y.-F., Lü, P., & Zhao, W. (2019) · International Conference on Multimedia Retrieval · doi:10.1145/3323873.3325018 · Cited by 15
“According to the ADC procedure [7], the computation is performed based on lookup tables.”
BackgroundMethodInfluentialDeep supervised hashing using symmetric relative entropy
Zhang, X., Zhou, L., Bai, X. P., Luan, X., Luo, J., & Hancock, E. R. (2019) · Pattern Recognition Letters · doi:10.1016/j.patrec.2019.07.010 · Cited by 15
Binary Code based Hash Embedding for Web-scale Applications
Yan, B., Wang, P., Liu, J., et al. (2021) · Proceedings of the ... ACM International Conference on Information & Knowledge Management · doi:10.1145/3459637.3482065 · Cited by 15
Return of the Lernaean Hydra: Experimental Evaluation of Data Series Approximate Similarity SearchPreprint
Echihabi, K., Zoumpatianos, K., Palpanas, T., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.11459 · Cited by 15
“In the past few years though, we are witnessing a renewed interest in the development of approximate methods [74, 156, 142, 18, 103].”
InfluentialGao, X., Ma, Z., Cheng, Y., et al. (2024) · Measurement · doi:10.1016/j.measurement.2024.114937 · Cited by 14
REIS: A High-Performance and Energy-Efficient Retrieval System with In-Storage Processing
Chen, K., Nadig, R., Frouzakis, M., Ghiasi, N. M., Liang, Y., Mao, H., Park, J., Sadrosadati, M., & Mutlu, O. (2025) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1145/3695053.3731116 · Cited by 20
“REIS offers improved performance against existing ISP-based ANNS accelerators, without introducing any hardware modifications, enabling easier adoption for RAG pipelines.”
Towards Engineering a Web-Scale Multimedia Service
Guðmundsson, G. Þ., Amsaleg, L., Jónsson, B. Þ., et al. (2017) · doi:10.1145/3083187.3083200 · Cited by 14
“proposed an indexing scheme based on the notion of product quantization and evaluated its performance by indexing 2 billion vectors [12].”
BackgroundEPQuant: A Graph Neural Network compression approach based on product quantization
Huang, L., Zhang, Z., Du, Z., et al. (2022) · Neurocomputing · doi:10.1016/j.neucom.2022.06.097 · Cited by 14
M5: Multi-Modal Multi-Interest Multi-Scenario Matching for Over-the-Top Recommendation
Zhao, P., Gao, X., Xu, C., & Chen, L. (2023) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3580305.3599863 · Cited by 16
“Facing the large item corpus in the web-scale system, ANN or MIPS [18, 19, 30] are often adopted to approximate the precise item retrieval during the online serving.”
MethodAsymmetric Feature Fusion for Image Retrieval
Wu, H., Wang, M., Zhou, W., et al. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.01066 · Cited by 14
“For efficient retrieval, a dynamic mixer is further proposed to aggregate diverse gallery features into compact embedding, which allows efficient vector search [11, 18] to be exploited.”
BackgroundConsistent penalizing field loss for zero-shot image retrieval
Liu, C., She, W., Chen, M., Li, X., & Yang, S. X. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.121287 · Cited by 14
Relative NN-Descent: A Fast Index Construction for Graph-Based Approximate Nearest Neighbor Search
Ono, N., & Matsui, Y. (2023) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3581783.3612290 · Cited by 13
Enhancing Generative Retrieval with Reinforcement Learning from Relevance Feedback
Zhou, Y., Dou, Z., & Wen, J.-R. (2023) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2023.emnlp-main.768 · Cited by 18
“These methods convert queries and documents into dense vectors and employ ANNS (Jégou et al., 2011) for efficient vector-based search.”
Yang, X., Li, S., Li, M., et al. (2023) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3341146 · Cited by 14
“Jégou et al. [14] proposed two heterogeneous signcryption schemes to ensure the security of heterogeneous V2V communication in VANET: one is that the vehicle in PKI sends messages to the vehicle in IBC; another solution is to swap the identities of the two.”
BackgroundExperimental Analysis of Large-Scale Learnable Vector Storage Compression
Zhang, H., Zhao, P., Miao, X., Shao, Y., Liu, Z., Yang, T., & Cui, B. (2023) · Proceedings of the VLDB Endowment · doi:10.14778/3636218.3636234 · Cited by 17
Improving Dysarthric Speech Segmentation With Emulated and Synthetic Augmentation
Naeini, S. A., Simmatis, L. E. R., Jafari, D., Yunusova, Y., & Taati, B. (2024) · IEEE Journal of Translational Engineering in Health and Medicine · doi:10.1109/jtehm.2024.3375323 · Cited by 17
“W2V is a framework for self-supervised training that can be broken down into feature encoder, contextualised representation, and quantisation module [36].”
Zhao, X., & Liu, J. (2024) · Knowledge-Based Systems · doi:10.1016/j.knosys.2024.112541 · Cited by 14
Zhan, Y., Yang, R., Zhang, Y., & Wang, Z. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3500047 · Cited by 15
“Efficient processing and storage of these features are achieved through PQ [15], a compression method highly suitable for storing and replaying encoded features.”
Wei, J., Lee, X., Liao, Z., et al. (2025) · Proceedings of the ACM on Management of Data · doi:10.1145/3709729 · Cited by 14
Xu, Q., Yang, J., Zhang, F., Pan, J., Chen, K., Shen, Y., Zhou, A. C., & Du, X. (2025) · Proceedings of the ACM on Management of Data · doi:10.1145/3709743 · Cited by 16
“There are mainly two types of indexes used: graph-based indexes and partition-based indexes (which include cluster-based [3, 5, 8, 17, 31– 33, 74, 77], hash-based [11, 28, 56, 66, 71], and high-dimensional tree-based [6, 42, 49, 64] methods).”
MGraph: multimodal event summarization in social media using topic models and graph-based ranking
Schinas, M., Papadopoulos, S., Kompatsiaris, I. Y., et al. (2015) · International Journal of Multimedia Information Retrieval · doi:10.1007/s13735-015-0089-9 · Cited by 14
“Fast retrieval of top-k temporal neighbours takes place using range queries on a B-Tree index, retrieval of top-k textual neighbours using an implementation of Locality Sensitive Hashing for the cosine similarity metric [8], and retrieval of top-k visual neighbours using Product Quantization [16].”
MethodCoordinated Local Metric Learning
Saxena, S., & Verbeek, J. J. (2015) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2015.56 · Cited by 14
“This means that the local metrics cannot be used for global data visualization, and do not support efficient retrieval techniques based on 2 -quantization, such as product quantization and multiple-assignment retrieval [21].”
BackgroundMethodInfluentialGlobal Hashing System for Fast Image Search
Tian, D., & Tao, D. (2016) · IEEE Transactions on Image Processing · doi:10.1109/tip.2016.2617081 · Cited by 14
“B ∈ {−1, 1}n×c, B>B = nI, B>1 = 0.”
Unsupervised deep hashing with stacked convolutional autoencoders
En, S., Crémilleux, B., & Jurie, F. (2017) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2017.8296917 · Cited by 13
“This is the reason why a new trend in hashing is to consider semi-supervised or even unsupervised learning strategies [1, 6, 12, 13].”
BackgroundDeep Clustering and Block Hashing Network for Face Image Retrieval
Jang, Y. K., Jeong, D.-J., Lee, S. H., & Cho, N. I. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-20876-9_21 · Cited by 13
Unsupervised Rank-Preserving Hashing for Large-Scale Image Retrieval
Karaman, S., Lin, X., Hu, X., et al. (2019) · doi:10.1145/3323873.3325038 · Cited by 14
Recurrent Binary Embedding for GPU-Enabled Exhaustive Retrieval from Billion-Scale Semantic Vectors
Shan, Y., Jiao, J., Zhu, J., et al. (2018) · ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (KDD) · doi:10.1145/3219819.3220018 · Cited by 14
Deep learning model with low-dimensional random projection for large-scale image search
Alzu’bi, A., & Abuarqoub, A. (2020) · Engineering Science and Technology an International Journal · doi:10.1016/j.jestch.2019.12.004 · Cited by 16
Efficient Object Embedding for Spliced Image Retrieval
Chen, B.-C., Wu, Z., Davis, L. S., & Lim, S.-N. (2021) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr46437.2021.01472 · Cited by 20
“Later advances on instance retrieval using local features [39] and indexing methods [62, 30, 31] achieved robustness against illumination and geometric variations.”
MethodLearning An End-to-End Structure for Retrieval in Large-Scale Recommendations
Gao, W., Fan, X., Wang, C., et al. (2021) · Proceedings of the ... ACM International Conference on Information & Knowledge Management · doi:10.1145/3459637.3482362 · Cited by 14
“Efficient MIPS or ANN algorithms include tree-based algorithms [11, 23], locality sensitive hashing (LSH) [27, 28], product quantization (PQ) [6, 13], hierarchical navi-gable small world graphs (HNSW) [18], etc.”
BackgroundJi, Z.-R., & Wang, C.-L. (2022) · International Conference on Supercomputing · doi:10.1145/3524059.3532368 · Cited by 14
Internal Similarity Network for Rejoining Oracle Bone Fragment Images
Zhang, Z., Guo, A., & Li, B. (2022) · Symmetry · doi:10.3390/sym14071464 · Cited by 14
Learning-based query optimization for multi-probe approximate nearest neighbor search
Zhang, P., Yao, B., Gao, C., Wu, B., He, X., Li, F., Lu, Y., Zhan, C., & Tang, F. (2022) · The VLDB Journal · doi:10.1007/s00778-022-00762-0 · Cited by 15
“Somewell-knownandwidelyused approaches, such as Locality Sensitive Hashing (LSH) [18] and Product Quantization (PQ) [27], all fall into the above categories.”
BackgroundMethodA Study of Speech Recognition for Kazakh Based on Unsupervised Pre-Training
Meng, W., & Yolwas, N. (2023) · Sensors · doi:10.3390/s23020870 · Cited by 14
“At the same time, the quantization module discretizes Z into a finite set of speech representations, using the method of product quantization [32] in the discretization process.”
Classification of complex biological aging images using fuzzy Kolmogorov–Sinai entropy
Pham, T. D. (2014) · Journal of Physics D Applied Physics · doi:10.1088/0022-3727/47/48/485402 · Cited by 14
“The method of VQ has been used for efficient nearest neighbor search [45].”
MethodMultimodal biomedical image indexing and retrieval using descriptive text and global feature mapping
Simpson, M. S., Demner‐Fushman, D., Antani, S., & Thoma, G. R. (2013) · Information Retrieval · doi:10.1007/s10791-013-9235-2 · Cited by 14
“GFM’s use of lower-dimensional feature vectors is related to the notion of product quantization (Jégou et al. 2011) and the dimensionality reduction technique proposed by Ferhatosmanoglu et al. (2001) that partitions vectors after having transformed them using the Karhunen–Loeve transformation…”
MethodA New ETL Approach Based on Data Virtualization
Guo, S.-S., Yuan, Z., Sun, A., et al. (2015) · Journal of Computer Science and Technology · doi:10.1007/s11390-015-1524-3 · Cited by 14
“Another type of relative approaches is the PQ method[18-19].”
Optimization of Robust Loss Functions for Weakly-Labeled Image Taxonomies
McAuley, J., Arnau Ramisa, A., & Caetano, T. S. (2012) · International Journal of Computer Vision · doi:10.1007/s11263-012-0561-4 · Cited by 15
“In their approach, they used high-dimensional Fisher Kernels for image representation with lossy compression techniques: first, dimensionality reduction using Hash Kernels (Shi et al. 2009) was attempted and secondly, since the results degraded rapidly with smaller descriptor dimensionality, coding with Product Quantiz…”
Xia, W., Mita, Y., & Shibata, T. (2015) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2015.2437901 · Cited by 14
The distributed system for inverted multi-index visual retrieval
Lin, X., Shen, Y., Cai, L., et al. (2016) · Neurocomputing · doi:10.1016/j.neucom.2015.11.131 · Cited by 12
“Firstly, the local features are extracted and quantized into visual words [1, 2, 4, 5] for both query and reference images.”
BackgroundMethodInfluentialEriksson, E., Dán, G., & Fodor, V. (2017) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2017.2745102 · Cited by 14
Learning a Complete Image Indexing Pipeline
Jain, H., Zepeda, J., Pérez, P., & Gribonval, R. (2018) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2018.00518 · Cited by 13
“Structured variants of VQ, on the other hand, produce fine-grain approximations of the features through very compact codes [3, 8, 9, 12, 15, 16, 20, 29] that enable look-up table-based efficient distance computations.”
BackgroundMethodInfluentialQuarter-Point Product Quantization for approximate nearest neighbor search
An, S., Huang, Z., Bai, S., Che, G., Ma, X., Luo, J., & Chen, Y. (2019) · Pattern Recognition Letters · doi:10.1016/j.patrec.2019.04.017 · Cited by 14
Online multimedia retrieval on CPU–GPU platforms with adaptive work partition
de Souza, R. M., Fernandes, A. G. D. A., Teixeira, T. S. F. X., Teodoro, G., & Ferreira, R. (2020) · Journal of Parallel and Distributed Computing · doi:10.1016/j.jpdc.2020.10.001 · Cited by 13
Sub-Selective Quantization for Large-Scale Image Search
Li, Y., Chen, C., Liu, W., & Huang, J. (2014) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v28i1.9138 · Cited by 13
“All rights reserved. amount of research in binary code generation such as (Andoni and Indyk 2006)(Jegou, Douze, and Schmid 2011)(Ge et al. 2013)(Raginsky and Lazebnik 2009)(Kulis and Grauman 2012) (Weiss, Torralba, and Fergus 2008)(Weiss, Fergus, and Torralba 2012)(Gong et al. 2013)(Liu et al.…”
BackgroundSampling-Decomposable Generative Adversarial RecommenderPreprint
Jin, B., Lian, D., Liu, Z., Liu, Q., Ma, J., Xie, X., & Chen, E. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2011.00956 · Cited by 24
“As discussed in this work, it is also more possible for the framework to seamlessly integrate approximate search algorithms, such as ALSH [28, 21], PQ [14] and HNSW [20], with complex recommendation algorithms, so as to mutually reinforce the learning of both algorithms.”
FreshDiskANN: A Fast and Accurate Graph-Based ANN Index for Streaming Similarity SearchPreprint
Singh, A., Subramanya, S. J., Krishnaswamy, R., & Simhadri, H. V. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2105.09613 · Cited by 138
“Popular methods like IVFADC [35], OPQ [29], LOPQ [39], FAISS [38], IVFOADC+G+P [12] and IMI [8] exploit the data distribution to produce low”
BackgroundMethodInfluentialResults of the NeurIPS'21 Challenge on Billion-Scale Approximate Nearest Neighbor SearchPreprint
Simhadri, H. V., Williams, G., Aumüller, M., Douze, M., Babenko, A., Baranchuk, D., Chen, Q., Hosseini, L., Krishnaswamy, R., Srinivasa, G., et al. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2205.03763 · Cited by 31
“, (Beygelzimer et al., 2006; Babenko and Lempitsky, 2014; Johnson et al., 2017; Weber et al., 1998; Baranchuk et al., 2018; Malkov and Yashunin, 2016; Jégou et al., 2011; Arya and Mount, 1993; Indyk and Motwani, 1998; Iwasaki and Miyazaki, 2018; Guo et al., 2020; Aumüller et al., 2019) for some recent works, and also…”
BackgroundMethodResultInfluentialSelf-citationJauhari, K., Rahman, A. Z., Huda, M. A., Widodo, A., & Prahasto, T. (2025) · Mechanical Systems and Signal Processing · doi:10.1016/j.ymssp.2025.112357 · Cited by 14
GaussDB-Vector: A Large-Scale Persistent Real-Time Vector Database for LLM Applications
Sun, J., Li, G., Pan, J., Wang, J., Xie, Y., Liu, R., & Nie, W. (2025) · Proceedings of the VLDB Endowment · doi:10.14778/3750601.3750619 · Cited by 13
“GaussDB-Vector adopts PQ-ADC [24] to increase the precision of product quantization, PQ-ADC only quantizes the data in the index, and the query vector remains unchanged.”
Adaptive Quantization for Hashing: An Information-Based Approach to Learning Binary Codes
Xiong, C., Chen, W., Chen, G., Johnson, D., & Corso, J. J. (2014) · SIAM International Conference on Data Mining · doi:10.1137/1.9781611973440.20 · Cited by 13
“Large-scale data mining and retrieval applications have increasingly turned to compact binary data representations as a way to achieve both fast queries and efficient data storage; many algorithms have been proposed for learning effective binary encodings.”
BackgroundLinear Distance Preserving Pseudo-Supervised and Unsupervised Hashing
Wang, M., Zhou, W., Tian, Q., Zha, Z.-J., & Li, H. (2016) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/2964284.2964334 · Cited by 13
“We randomly select 10,000 pairs of data points from ANN SIFT1M dataset [8] and compute their Euclidean distances.”
BackgroundMethodInfluentialJeǵou, H. (2017) · doi:10.1145/3122865.3122871 · Cited by 13
Multi-modal Dictionary BERT for Cross-modal Video Search in Baidu Advertising
Yu, T., Yang, Y., Li, Y., Liu, L., Sun, M., & Li, P. (2021) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3459637.3481937 · Cited by 13
“To alleviate the computation andmemory burden, we propose to utilize product quantization (PQ) [14, 23] to map each feature into a Cartesian product of several sub-codewords.”
MethodSEAnet: A Deep Learning Architecture for Data Series Similarity Search
Wang, Q., & Palpanas, T. (2023) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2023.3270264 · Cited by 13
“…incremental learning techniques for quickly fitting new or dynamic datasets, the development of more powerful sampling strategies, and the careful study of query answering strategies on top of DEA, including product quantization [48], locality sensitive hashing [49], and modern series indexes [30].”
Staleness-Reduction Mini-Batch K-Means
Zhu, X., Sun, J., He, Z., et al. (2023) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2023.3279122 · Cited by 13
“2) Dataset: We run our experiments with four real-world datasets: Sift [7], Gist [7], Poker [8], and Mnist8m [9], as shown in Table II.”
MethodLearning Balanced Tree Indexes for Large-Scale Vector Retrieval
Li, W., Feng, C., Lian, D., Xie, Y., Liu, H., Ge, Y., & Chen, E. (2023) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3580305.3599406 · Cited by 13
Thinking Fast and Slow: Efficient Text-to-Visual Retrieval with Transformers
Miech, A., Alayrac, J.-B., Laptev, I., Sivic, J., & Zisserman, A. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.00970 · Cited by 149
Optimizing Dense Retrieval Model Training with Hard Negatives
Zhan, J., Mao, J., Liu, Y., Guo, J., Zhang, M., & Ma, S. (2021) · arXiv · doi:10.48550/arxiv.2104.08051 · Cited by 251
Metadata
- Authors
- Jégou, H., Douze, M., & Schmid, C.
- Affiliations
- Institut national de recherche en sciences et technologies du numérique (FR)
- Venue
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- Published
- 2011-01
- Volume
- 33
- Issue
- 1
- Pages
- 117-128
- ISSN
- 0162-8828 (ISSN-L)
- OpenAlex
- W2124509324
- Semantic Scholar
- 4748d22348e72e6e06c2476486afddbc76e5eca7
- Cited by
- 2,956
- References
- 35
- Citations by year
- 2026: 260 · 2025: 295 · 2024: 233 · 2023: 211 · 2022: 195 · 2021: 204 · 2020: 187 · 2019: 207 · 2018: 201 · 2017: 184 · 2016: 196 · 2015: 196 · 2014: 159 · 2013: 116 · 2012: 71
- FWCI
- 84.24
- Type
- journal article
- Language
- en
- Open access
- green
- inria.hal.science
- Locations
- IEEE Transactions on Pattern Analysis and Machine Intelligence (publishedVersion)
- PubMed (publishedVersion)
- HAL (submittedVersion, other-oa)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image and Video Retrieval Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Retrieval and Classification Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Data Compression Techniques
- Keywords
- product quantization, approximate nearest neighbor search, subspace decomposition, inverted index, SIFT descriptors, GIST descriptors, image retrieval, scalability
- MeSH terms
- Algorithms, Artificial Intelligence, Image Interpretation, Computer-Assisted, Image Processing, Computer-Assisted, Pattern Recognition, Automated, Models, Statistical, Cluster Analysis, Information Storage and Retrieval
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A second model labelled each claim on 2026-10-08 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
