ABSTRACT
Hashing, which represents data items as compact binary codes, has been becoming a more and more popular technique, e.g., for large-scale image retrieval, owing to its super fast search speed as well as its extremely economical memory consumption. However, existing hashing methods all try to learn binary codes from artificially balanced datasets which are not commonly available in real-world scenarios. In this paper, we propose Long-Tail Hashing Network (LTHNet), a novel two-stage deep hashing approach that addresses the problem of learning to hash for more realistic datasets where the data labels roughly exhibit a long-tail distribution. Specifically, the first stage is to learn relaxed embeddings of the given dataset with its long-tail characteristic taken into account via an end-to-end deep neural network; the second stage is to binarize those obtained embeddings. A critical part of LTHNet is its dynamic meta-embedding module extended with a determinantal point process which can adaptively realize visual knowledge transfer between head and tail classes, and thus enrich image representations for hashing. Our experiments have shown that LTHNet achieves dramatic performance improvements over all state-of-the-art competitors on long-tail datasets, with no or little sacrifice on balanced datasets. Further analyses reveal that while to our surprise directly manipulating class weights in the loss function has little effect, the extended dynamic meta-embedding module, the usage of cross-entropy loss instead of square loss, and the relatively small batch-size for training all contribute to LTHNet's success.
Supplemental Material
- Reda Alhajj and Jon G. Rokne. 2018. Learning to Rank .Encyclopedia of Social Network Analysis and Mining, 2nd Edition, Springer.Google Scholar
- Zhangjie Cao, Mingsheng Long, Jianmin Wang, and Philip S. Yu. 2017. HashNet: Deep Learning to Hash by Continuation. In ICCV. 5609--5618.Google Scholar
- Suthee Chaidaroon, Travis Ebesu, and Yi Fang. 2018. Deep Semantic Text Hashing with Weak Supervision. In SIGIR. ACM , 1109--1112.Google Scholar
- Nitesh V. Chawla, Kevin W. Bowyer, Lawrence O. Hall, and W. Philip Kegelmeyer. 2002. SMOTE: Synthetic Minority Over-sampling Technique. JAIR , Vol. 16 (2002), 321--357.Google ScholarCross Ref
- Laming Chen, Guoxin Zhang, and Eric Zhou. 2018. Fast Greedy MAP Inference for Determinantal Point Process to Improve Recommendation Diversity. In NeurIPS. 5627--5638.Google Scholar
- Yong Chen, Zhibao Tian, Hui Zhang, Jun Wang, and Dell Zhang. 2020 a. Strongly Constrained Discrete Hashing. TIP , Vol. 29 (2020), 3596--3611.Google Scholar
- Yong Chen, Hui Zhang, Zhibao Tian, Jun Wang, Dell Zhang, and Xuelong Li. 2020 c. Enhanced Discrete Multi-modal Hashing: More Constraints yet Less Time to Learn. IEEE TKDE (2020), 1--13.Google Scholar
- Zhihong Chen, Rong Xiao, Chenliang Li, Gangfeng Ye, Haochuan Sun, and Hongbo Deng. 2020 b. ESAM: Discriminative Domain Adaptation with Non-Displayed Items to Improve Long-Tail Performance. In SIGIR. 579--588.Google Scholar
- Aaron Clauset, Cosma Rohilla Shalizi, and Mark EJ Newman. 2009. Power-Law Distributions in Empirical Data. SIAM Rev. , Vol. 51, 4 (2009), 661--703.Google ScholarDigital Library
- Yin Cui, Menglin Jia, Tsung-Yi Lin, Yang Song, and Serge J. Belongie. 2019. Class-Balanced Loss Based on Effective Number of Samples. In CVPR. 9268--9277.Google Scholar
- Yin Cui, Yang Song, Chen Sun, Andrew Howard, and Serge J. Belongie. 2018. Large Scale Fine-Grained Categorization and Domain-Specific Transfer Learning. In CVPR . 4109--4118.Google Scholar
- Cheng Deng, Zhaojia Chen, Xianglong Liu, Xinbo Gao, and Dacheng Tao. 2018. Triplet-Based Deep Hashing Network for Cross-Modal Retrieval. TIP , Vol. 27, 8 (2018), 3893--3903.Google ScholarCross Ref
- Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Fei-Fei Li. 2009. ImageNet: A Large-Scale Hierarchical Image Database. In CVPR. 248--255.Google Scholar
- Doug Downey, Susan T. Dumais, and Eric Horvitz. 2007. Heads and Tails: Studies of Web Search with Common and Rare Queries. In SIGIR . 847--848.Google Scholar
- Norbert Fuhr. 2018. Some Common Mistakes in IR Evaluation, and How They Can Be Avoided. , Vol. 51, 3 (2018), 32--41.Google ScholarDigital Library
- Lu Gan, Diana Nurbakova, Lé a Laporte, and Sylvie Calabretto. 2020. Enhancing Recommendation Diversity using Determinantal Point Processes on Knowledge Graphs. In SIGIR. 2001--2004.Google Scholar
- Ming Gao, Leihui Chen, Xiangnan He, and Aoying Zhou. 2018. BiNE: Bipartite Network Embedding. In SIGIR. 715--724.Google Scholar
- Dar'io Garigliotti, Dyaa Albakour, Miguel Martinez, and Krisztian Balog. 2019. Unsupervised Context Retrieval for Long-tail Entities. In SIGIR. 225--228.Google Scholar
- Aristides Gionis, Piotr Indyk, and Rajeev Motwani. 1999. Similarity Search in High Dimensions via Hashing. In VLDB. 518--529.Google Scholar
- Yunchao Gong and Svetlana Lazebnik. 2011. Iterative Quantization: A Procrustean Approach to Learning Binary Codes. In CVPR . 817--824.Google Scholar
- Yunchao Gong, Svetlana Lazebnik, Albert Gordo, and Florent Perronnin. 2013. Iterative Quantization: A Procrustean Approach to Learning Binary Codes for Large-Scale Image Retrieval. IEEE TPAMI , Vol. 35, 12 (2013), 2916--2929.Google ScholarDigital Library
- Jie Gui and Ping Li. 2018. R2SDH: Robust Rotated Supervised Discrete Hashing. In KDD. 1485--1493.Google Scholar
- Jie Gui, Tongliang Liu, Zhenan Sun, Dacheng Tao, and Tieniu Tan. 2018. Fast Supervised Discrete Hashing. IEEE TPAMI , Vol. 40, 2 (2018), 490--496.Google ScholarDigital Library
- Hui Han, Wenyuan Wang, and Binghuan Mao. 2005. Borderline-SMOTE: A New Over-Sampling Method in Imbalanced Data Sets Learning. In ICIC . 878--887.Google Scholar
- Casper Hansen, Christian Hansen, Jakob Grue Simonsen, Stephen Alstrup, and Christina Lioma. 2019. Unsupervised Neural Generative Semantic Hashing. In SIGIR. 735--744.Google Scholar
- Casper Hansen, Christian Hansen, Jakob Grue Simonsen, Stephen Alstrup, and Christina Lioma. 2020 a. Content-aware Neural Hashing for Cold-start Recommendation. In SIGIR . 971--980.Google Scholar
- Casper Hansen, Christian Hansen, Jakob Grue Simonsen, Stephen Alstrup, and Christina Lioma. 2020 b. Unsupervised Semantic Hashing with Pairwise Reconstruction. In SIGIR . 2009--2012.Google Scholar
- Haibo He and Edwardo A. Garcia. 2009. Learning from Imbalanced Data. IEEE TKDE , Vol. 21 (2009), 1263--1284.Google Scholar
- Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. Deep Residual Learning for Image Recognition. In CVPR. 770--778.Google Scholar
- Xiangyu He, Peisong Wang, and Jian Cheng. 2019. K-Nearest Neighbors Hashing. In CVPR. 2839--2848.Google Scholar
- Yen-Chang Hsu, Zhaoyang Lv, and Zsolt Kira. 2018. Learning to Cluster in order to Transfer Across Domains and Tasks. In ICLR . 1--20.Google Scholar
- Bingyi Kang, Saining Xie, Marcus Rohrbach, Zhicheng Yan, Albert Gordo, Jiashi Feng, and Yannis Kalantidis. 2020. Decoupling Representation and Classifier for Long-Tailed Recognition. In ICLR . 1--16.Google Scholar
- Qi Kang, Xiaoshuang Chen, Sisi Li, and MengChu Zhou. 2017. A Noise-Filtered Under-Sampling Scheme for Imbalanced Classification. IEEE Transactions on Cybernetics , Vol. 47, 12 (2017), 4263--4274.Google ScholarCross Ref
- Wang-Cheng Kang, Wu-Jun Li, and Zhi-Hua Zhou. 2016. Column Sampling based Discrete Supervised Hashing. In AAAI. 1230--1236.Google Scholar
- Gou Koutaki, Keiichiro Shirai, and Mitsuru Ambai. 2018. Hadamard Coding for Supervised Discrete Hashing. TIP , Vol. 27, 11 (2018), 5378--5392.Google ScholarDigital Library
- Alex Krizhevsky. 2009. Learning Multiple Layers of Features from Tiny Images . Technical Report. University of Toronto. 1----60 pages.Google Scholar
- Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton. 2012. ImageNet Classification with Deep Convolutional Neural Networks. In NeurIPS . 1106--1114.Google ScholarDigital Library
- Qi Li, Zhenan Sun, Ran He, and Tieniu Tan. 2017. Deep Supervised Discrete Hashing. In NeurIPS. 2482--2491.Google Scholar
- Qi Li, Zhenan Sun, Ran He, and Tieniu Tan. 2020. A General Framework for Deep Supervised Discrete Hashing. IJCV , Vol. 128, 8 (2020), 2204--2222.Google ScholarDigital Library
- Wu-Jun Li, Sheng Wang, and Wang-Cheng Kang. 2016. Feature Learning Based Deep Supervised Hashing with Pairwise Labels. In IJCAIs . 1711--1717.Google Scholar
- Guohua Liang and Chengqi Zhang. 2012. An efficient and simple under-sampling technique for imbalanced time series classification. In CIKM. 2339--2342.Google Scholar
- Guosheng Lin, Chunhua Shen, Qinfeng Shi, Anton van den Hengel, and David Suter. 2014. Fast Supervised Hashing with Decision Trees for High-Dimensional Data. In CVPR . 1971--1978.Google Scholar
- Kevin Lin, Jiwen Lu, Chu-Song Chen, and Jie Zhou. 2016. Learning Compact Binary Descriptors with Unsupervised Deep Neural Networks. In CVPR . 1183--1192.Google Scholar
- Tsung-Yi Lin, Priya Goyal, Ross B. Girshick, Kaiming He, and Piotr Dollá r. 2017. Focal Loss for Dense Object Detection. In ICCV. 2999--3007.Google Scholar
- Tsung-Yi Lin, Priya Goyal, Ross B. Girshick, Kaiming He, and Piotr Dollá r. 2020. Focal Loss for Dense Object Detection. IEEE TPAMI , Vol. 42, 2 (2020), 318--327.Google ScholarCross Ref
- Jack Lindsey, Samuel A. Ocko, Surya Ganguli, and Sté phane Deny. 2019. A Unified Theory of Early Visual Representations from Retina to Cortex through Anatomically Constrained Deep CNNs. In ICLR. 1--17.Google Scholar
- Jingzhou Liu, Wei-Cheng Chang, Yuexin Wu, and Yiming Yang. 2017. Deep Learning for Extreme Multi-Label Text Classification. In SIGIR . 115--124.Google Scholar
- Song Liu, Shengsheng Qian, Yang Guan, Jiawei Zhan, and Long Ying. 2020. Joint-modal Distribution-based Similarity Hashing for Large-scale Unsupervised Deep Cross-modal Retrieval. In SIGIR. 1379--1388.Google Scholar
- Tie-Yan Liu. 2009. Learning to Rank for Information Retrieval .Foundations and Trends in Information Retrieval.Google Scholar
- Wei Liu, Cun Mu, Sanjiv Kumar, and Shih-Fu Chang. 2014. Discrete Graph Hashing. In NeurIPS. 3419--3427.Google Scholar
- Wei Liu, Jun Wang, Rongrong Ji, Yu-Gang Jiang, and Shih-Fu Chang. 2012. Supervised hashing with kernels. In CVPR. 2074--2081.Google Scholar
- Ziwei Liu, Zhongqi Miao, Xiaohang Zhan, Jiayun Wang, Boqing Gong, and Stella X. Yu. 2019. Large-Scale Long-Tailed Recognition in an Open World. In CVPR. 2537--2546.Google Scholar
- Fuchen Long, Ting Yao, Qi Dai, Xinmei Tian, Jiebo Luo, and Tao Mei. 2018. Deep Domain Adaptation Hashing with Adversarial Learning. In SIGIR. 725--734.Google Scholar
- Xu Lu, Lei Zhu, Zhiyong Cheng, Liqiang Nie, and Huaxiang Zhang. 2019. Online Multi-modal Hashing with Dynamic Query-adaption. In SIGIR . 715--724.Google Scholar
- Xu Lu, Lei Zhu, Jingjing Li, Huaxiang Zhang, and Heng Tao Shen. 2020. Efficient Supervised Discrete Multi-View Hashing for Large-Scale Multimedia Search. TMM , Vol. 22, 8 (2020), 2048--2060.Google ScholarCross Ref
- Xin Luo, Liqiang Nie, Xiangnan He, Ye Wu, Zhen-Duo Chen, and Xin-Shun Xu. 2018. Fast Scalable Supervised Hashing. In SIGIR. 735--744.Google Scholar
- Mark EJ Newman. 2005. Power Laws, Pareto Distributions and Zipf's Law. Contemporary Physics , Vol. 46, 5 (2005), 323--351.Google ScholarCross Ref
- Casper Petersen, Jakob Grue Simonsen, and Christina Lioma. 2016. Power Law Distributions in Information Retrieval. ACM Transactions on Information Systems (TOIS) , Vol. 34, 2 (2016), 8:1--8:37.Google ScholarDigital Library
- William J. Reed. 2001. The Pareto, Zipf and other Power Laws. Economics Letters , Vol. 74, 1 (2001), 15--19.Google ScholarCross Ref
- David E. Rumelhart, Geoffrey E. Hinton, and Ronald J. Williams. 1986. Learning Representations by Back-Propagating Errors. Nature , Vol. 323 (1986), 533----536.Google ScholarCross Ref
- Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, Sanjeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael S. Bernstein, Alexander C. Berg, and Fei-Fei Li. 2015. ImageNet Large Scale Visual Recognition Challenge. IJCV , Vol. 115, 3 (2015), 211--252.Google ScholarDigital Library
- Tetsuya Sakai. 2020. On Fuhr's Guideline for IR Evaluation. , Vol. 54, 1 (2020), p14.Google Scholar
- Adam Santoro, Sergey Bartunov, Matthew Botvinick, Daan Wierstra, and Timothy P. Lillicrap. 2016. Meta-Learning with Memory-Augmented Neural Networks. In ICML. 1842--1850.Google ScholarDigital Library
- Fumin Shen, Chunhua Shen, Wei Liu, and Heng Tao Shen. 2015. Supervised Discrete Hashing. In CVPR. 37--45.Google Scholar
- Shaoyun Shi, Weizhi Ma, Min Zhang, Yongfeng Zhang, Xinxing Yu, Houzhi Shan, Yiqun Liu, and Shaoping Ma. 2020. Beyond User Embedding Matrix: Learning to Hash for Modeling Large-Scale Users in Recommendation. In SIGIR. 319--328.Google Scholar
- Karen Simonyan and Andrew Zisserman. 2015. Very Deep Convolutional Networks for Large-Scale Image Recognition. In ICLR . 1--14.Google Scholar
- Changchang Sun, Xuemeng Song, Fuli Feng, Wayne Xin Zhao, Hao Zhang, and Liqiang Nie. 2019. Supervised Hierarchical Cross-Modal Hashing. In SIGIR. 725--734.Google Scholar
- Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is All you Need. In NeurIPS. 5998--6008.Google Scholar
- Di Wang, Quan Wang, Yaqiang An, Xinbo Gao, and Yumin Tian. 2020 b. Online Collective Matrix Factorization Hashing for Large-Scale Cross-Media Retrieval. In SIGIR . 1409--1418.Google Scholar
- Jun Wang, Wei Liu, Andy X. Sun, and Yu-Gang Jiang. 2013a. Learning Hash Codes with Listwise Supervision. In ICCV. 3032--3039.Google Scholar
- Jingdong Wang, Ting Zhang, Jingkuan Song, Nicu Sebe, and Heng Tao Shen. 2018. A Survey on Learning to Hash. IEEE TPAMI , Vol. 40, 4 (2018), 769--790.Google ScholarCross Ref
- Qifan Wang, Luo Si, Zhiwei Zhang, and Ning Zhang. 2014. Active Hashing with Joint Data Example and Tag Selection. In SIGIR. 405--414.Google Scholar
- Qifan Wang, Dan Zhang, and Luo Si. 2013b. Semantic Hashing Using Tags and Topic Modeling. In SIGIR. 213--222.Google Scholar
- Qifan Wang, Zhiwei Zhang, and Luo Si. 2015. Ranking Preserving Hashing for Fast Similarity Search. In IJCAI. 3911--3917.Google Scholar
- Xudong Wang, Long Lian, Zhongqi Miao, Ziwei Liu, and Stella X. Yu. 2020 a. Long-ailed Recognition by Routing Diverse Distribution-Aware Experts. arXiv , Vol. arXiv:2010.01809 (2020), 1--14.Google Scholar
- Xiaofang Wang, Yi Shi, and Kris M. Kitani. 2016. Deep Supervised Hashing with Triplet Labels. In ACCV. 70--84.Google Scholar
- Yu-Xiong Wang, Deva Ramanan, and Martial Hebert. 2017. Learning to Model the Tail. In NeurIPS. 7029--7039.Google Scholar
- Zijian Wang, Zheng Zhang, Yadan Luo, and Zi Huang. 2019. Deep Collaborative Discrete Hashing with Semantic-Invariant Structure. In SIGIR . 905--908.Google Scholar
- Erkun Yang, Cheng Deng, Tongliang Liu, Wei Liu, and Dacheng Tao. 2018. Semantic Structure-based Unsupervised Deep Hashing. In IJCAI. 1064--1070.Google Scholar
- Zhan Yang, Jun Long, Lei Zhu, and Wenti Huang. 2020. Nonlinear Robust Discrete Hashing for Cross-Modal Retrieval. In SIGIR . 1349--1358.Google Scholar
- Li Yuan, Tao Wang, Xiaopeng Zhang, Francis E. H. Tay, Zequn Jie, Wei Liu, and Jiashi Feng. 2020. Central Similarity Quantization for Efficient Image and Video Retrieval. In CVPR . 3080--3089.Google Scholar
- Weixin Zeng, Xiang Zhao, Wei Wang, Jiuyang Tang, and Zhen Tan. 2020. Degree-Aware Alignment for Entities in Tail. In SIGIR. 811--820.Google Scholar
- Dan Zhang, Fei Wang, and Luo Si. 2011. Composite Hashing with Multiple Information Sources. In SIGIR. 225--234.Google Scholar
- Dell Zhang, Jun Wang, Deng Cai, and Jinsong Lu. 2010. Self-Taught Hashing for Fast Similarity Search. In SIGIR. 18--25.Google Scholar
- Hongfei Zhang, Xia Song, Chenyan Xiong, Corby Rosset, Paul N Bennett, Nick Craswell, and Saurabh Tiwary. 2019. Generic Intent Representation in Web Search. In SIGIR. 65--74.Google Scholar
- Peichao Zhang, Wei Zhang, Wu-Jun Li, and Minyi Guo. 2014b. Supervised Hashing with Latent Factor Models. In SIGIR. 173--182.Google Scholar
- Xiao Zhang, Zhiyuan Fang, Yandong Wen, Zhifeng Li, and Yu Qiao. 2017. Range Loss for Deep Face Recognition with Long-Tailed Training Data. In ICCV . 5419--5428.Google Scholar
- Zhiwei Zhang, Qifan Wang, Lingyun Ruan, and Luo Si. 2014a. Preference Preserving Hashing for Efficient Recommendation. In SIGIR . 183--192.Google Scholar
- Boyan Zhou, Quan Cui, Xiu-Shen Wei, and Zhao-Min Chen. 2020. BBN: Bilateral-Branch Network With Cumulative Learning for Long-Tailed Visual Recognition. In CVPR. 9716--9725.Google Scholar
- Linchao Zhu and Yi Yang. 2020. Inflated Episodic Memory With Region Self-Attention for Long-Tailed Visual Recognition. In CVPR . 4343--4352.Google Scholar
Index Terms
- Long-Tail Hashing
Recommendations
Order preserving hashing for approximate nearest neighbor search
MM '13: Proceedings of the 21st ACM international conference on MultimediaIn this paper, we propose a novel method to learn similarity-preserving hash functions for approximate nearest neighbor (NN) search. The key idea is to learn hash functions by maximizing the alignment between the similarity orders computed from the ...
A Survey on Deep Hashing Methods
Nearest neighbor search aims at obtaining the samples in the database with the smallest distances from them to the queries, which is a basic task in a range of fields, including computer vision and data mining. Hashing is one of the most widely used ...
Linear Distance Preserving Pseudo-Supervised and Unsupervised Hashing
MM '16: Proceedings of the 24th ACM international conference on MultimediaWith the advantage in compact representation and efficient comparison, binary hashing has been extensively investigated for approximate nearest neighbor search. In this paper, we propose a novel and general hashing framework, which simultaneously ...
Comments