{"type":"ScholarlyArticle","title":"Deep Polarized Network for Supervised Learning of Accurate Binary Hashing Codes","authors":[{"name":"Lixin Fan","slug":"lixin-fan","given":"Lixin","family":"Fan","affiliation":"WeBank AI , Shenzhen , China"},{"name":"Kam Woh Ng","slug":"kam-woh-ng","given":"Kam Woh","family":"Ng","affiliation":"WeBank AI , Shenzhen , China"},{"name":"Ce Ju","slug":"ce-ju","given":"Ce","family":"Ju","affiliation":"WeBank AI , Shenzhen , China"},{"name":"Tianyu Zhang","slug":"tianyu-zhang","given":"Tianyu","family":"Zhang","affiliation":"WeBank AI , Shenzhen , China"},{"name":"Chee Seng Chan","slug":"chee-seng-chan","given":"Chee Seng","family":"Chan","affiliation":"University of Malaya , Kuala Lumpur , Malaysia"}],"abstract":"This paper proposes a novel deep polarized network (DPN) for learning to hash, in which each channel in the network outputs is  pushed far away from zero by employing a differentiable bit-wise hinge-like loss which is dubbed as polarization loss. Reformulated within a generic Hamming Distance Metric Learning framework [Norouzi et al., 2012], the proposed polarization loss bypasses the requirement to prepare pairwise labels for  (dis-)similar items and, yet, the proposed loss strictly bounds from above the pairwise Hamming Distance based losses.  The intrinsic connection between pairwise and pointwise label information, as disclosed in this paper, brings about the following methodological improvements: (a) we may directly employ the proposed differentiable polarization loss with no large deviations incurred from the target Hamming distance based loss;  and (b) the subtask of assigning binary codes becomes extremely simple --- even random codes assigned to  each class suffice to result in state-of-the-art performances, as demonstrated in CIFAR10, NUS-WIDE and ImageNet100 datasets.","doi":"10.24963/ijcai.2020/115","url":"https://ijcai.org/proceedings/2020/115","pdf":"https://ijcai.org/proceedings/2020/0115.pdf","markdown":"https://ijcai.org/proceedings/2020/115.md","bibtex":"https://ijcai.org/proceedings/2020/bibtex/115","volume":{"slug":"2020","title":"Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2020","url":"https://ijcai.org/proceedings/2020/","year":2020,"location":"Yokohama","dates":"January 2021 (postponed from 11-17 July 2020)","editors":"Christian Bessiere","isbn":"978-0-9992411-6-5","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2020"},"section":"Computer Vision","track":"Main track","number":115,"pages":{"first":825,"last":831},"keywords":["Data Mining: Big Data, Large-Scale Systems","Machine Learning: Clustering","Computer Vision: Big Data and Large Scale Methods"],"videos":[],"references":[{"position":1,"raw":"[Andoni and Indyk, 2006] A. Andoni and P. Indyk. Near-optimal hashing algorithms for approximate nearest neighbor in high di- mensions. In FOCS, pages 459-468, 2006.","title":"Near-optimal hashing algorithms for approximate nearest neighbor in high dimensions","authors":"A Andoni, P Indyk","year":"2006","venue":"FOCS","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Cakir et al., 2018] F. Cakir, K. He, and S. Sclaroff. Hashing with binary matrix pursuit. In ECCV, pages 344-361, 2018.","title":"Hashing with binary matrix pursuit","authors":"F Cakir, K He, S Sclaroff","year":"2018","venue":"ECCV","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Cakir et al., 2019] Fatih Cakir, Kun He, Sarah Bargal, and Stan Sclaroff. Hashing with mutual information. T-PAMI, 41(10):2424-2437, 2019.","title":"Hashing with mutual information","authors":"Fatih Cakir, Kun He, Sarah Bargal, Stan Sclaroff","year":"2019","venue":"T-PAMI","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Cao et al., 2017] Z. Cao, M. Long, J. Wang, and P. Yu. Hashnet: Deep learning to hash by continuation. In ICCV, 2017.","title":"Hashnet: Deep learning to hash by continuation","authors":"Z Cao, M Long, J Wang, P Yu","year":"2017","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Cao et al., 2018] Y. Cao, M. Long, L. Bin, and J. Wang. Deep cauchy hashing for hamming space retrieval. In CVPR, 2018.","title":"Deep cauchy hashing for hamming space retrieval","authors":"Y Cao, M Long, L Bin, J Wang","year":"2018","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Chua et al., 2009] T-S. Chua, J. Tang, R. Hong, H. Li, Z. Luo, and Y-T. Zheng. Nus-wide: A real-world web image database from national university of singapore. In CIVR, 2009.","title":"Nus-wide: A real-world web image database from national university of singapore","authors":"T-S Chua, J Tang, R Hong, H Li, Z Luo, Y-T Zheng","year":"2009","venue":"CIVR","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Fan et al., 2019] L. Fan, K. W. Ng, and C. S. Chan. Rethinking deep neural network ownership verification: Embedding pass- ports to defeat ambiguity attacks. In NeurIPS, pages 4716-4725, 2019.","title":"Rethinking deep neural network ownership verification: Embedding passports to defeat ambiguity attacks","authors":"L Fan, K W Ng, C S Chan","year":"2019","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Gong and Lazebnik, 2011] Y. Gong and S. Lazebnik. Iterative quantization: A procrustean approach to learning binary codes. In CVPR, 2011.","title":"Iterative quantization: A procrustean approach to learning binary codes","authors":"Y Gong, S Lazebnik","year":"2011","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[He et al., 2018] K. He, F. Cakir, S. A. Bargal, and S. Sclaroff. Hashing as tie-aware learning to rank. In CVPR, pages 4023- 4032, 2018.","title":"Hashing as tie-aware learning to rank","authors":"K He, F Cakir, S A Bargal, S Sclaroff","year":"2018","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Hu et al., 2018] Q. Hu, P. Wang, and J. Cheng. From hashing to cnns: Training binary weight networks via hashing. In AAAI, 2018.","title":"From hashing to cnns: Training binary weight networks via hashing","authors":"Q Hu, P Wang, J Cheng","year":"2018","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Indyk and Motwani, 1998] P. Indyk and R. Motwani. Approxi- mate nearest neighbors: Towards removing the curse of dimen- sionality. In ACM-STOC, pages 604-613, 1998.","title":"Approximate nearest neighbors: Towards removing the curse of dimensionality","authors":"P Indyk, R Motwani","year":"1998","venue":"ACM-STOC","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Jeong and Song, 2018] Y. Jeong and H. O. Song. Efficient end-to- end learning for quantizable representations. In ICML, 2018.","title":"Efficient end-toend learning for quantizable representations","authors":"Y Jeong, H O Song","year":"2018","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Krizhevsky et al., 2012] A. Krizhevsky, I. Sutskever, and G. Hin- ton. Imagenet classification with deep convolutional neural net- works. In NIPS, pages 1097-1105, 2012.","title":"Imagenet classification with deep convolutional neural networks","authors":"A Krizhevsky, I Sutskever, G Hinton","year":"2012","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Krizhevsky, 2009] A. Krizhevsky. Learning multiple layers of fea- tures from tiny images. Technical report, 2009.","title":"Learning multiple layers of features from tiny images","authors":"A Krizhevsky","year":"2009","venue":null,"doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Kulis and Darrell, 2009] B. Kulis and T. Darrell. Learning to hash with binary reconstructive embeddings. In NIPS, pages 1042- 1050, 2009.","title":"Learning to hash with binary reconstructive embeddings","authors":"B Kulis, T Darrell","year":"2009","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Li et al., 2016] W-J Li, S. Wang, and W-C Kang. Feature learning based deep supervised hashing with pairwise labels. In IJCAI, pages 1711-1717, 2016.","title":"Feature learning based deep supervised hashing with pairwise labels","authors":"W-J Li, S Wang, W-C Kang","year":"2016","venue":"IJCAI","doi":null,"url":null,"internal_url":"/proceedings/2016/245"},{"position":17,"raw":"[Lin et al., 2013] G. Lin, C. Shen, D. Suter, and A. van den Hengel. A general two-step approach to learning-based hashing. In ICCV, 2013.","title":"A general two-step approach to learning-based hashing","authors":"G Lin, C Shen, D Suter, A Van Den, Hengel","year":"2013","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Lin et al., 2015] K. Lin, H-F Yang, J-H Hsiao, and C-S Chen. Deep learning of binary hash codes for fast image retrieval. In CVPR-W, 2015.","title":"Deep learning of binary hash codes for fast image retrieval","authors":"K Lin, H-F Yang, J-H Hsiao, C-S Chen","year":"2015","venue":"CVPR-W","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Lin et al., 2017] G. Lin, F. Liu, C. Shen, J. Wu, and H. T. Shen. Structured learning of binary codes with column generation for optimizing ranking measures. Int. J. Comput. Vision, 123(2):287- 308, June 2017.","title":"Structured learning of binary codes with column generation for optimizing ranking measures","authors":"G Lin, F Liu, C Shen, J Wu, H T Shen","year":"2017","venue":"Int. J. Comput. Vision","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Liu et al., 2012] W. Liu, J. Wang, R. Ji, Y. Jiang, and S. Chang. Supervised hashing with kernels. In CVPR, pages 2074-2081, 2012.","title":"Supervised hashing with kernels","authors":"W Liu, J Wang, R Ji, Y Jiang, S Chang","year":"2012","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Liu et al., 2016] H. Liu, R. Wang, S. Shan, and X. Chen. Deep su- pervised hashing for fast image retrieval. In CVPR, pages 2064- 2072, 2016.","title":"Deep supervised hashing for fast image retrieval","authors":"H Liu, R Wang, S Shan, X Chen","year":"2016","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Liu et al., 2018] B. Liu, Y. Cao, M. Long, J. Wang, and J. Wang. Deep triplet quantization. In ACM-MM, pages 755-763, 2018.","title":"Deep triplet quantization","authors":"B Liu, Y Cao, M Long, J Wang, J Wang","year":"2018","venue":"ACM-MM","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Norouzi and Blei, 2011] M. Norouzi and D. M. Blei. Minimal loss hashing for compact binary codes. In ICML, pages 353-360, 2011.","title":"Minimal loss hashing for compact binary codes","authors":"M Norouzi, D M Blei","year":"2011","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Norouzi et al., 2012] M. Norouzi, D. Fleet, and R. Salakhutdinov. Hamming distance metric learning. In NIPS, 2012.","title":"Hamming distance metric learning","authors":"M Norouzi, D Fleet, R Salakhutdinov","year":"2012","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Paszke et al., 2019] A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, A. Desmaison, A. Köpf, E. Yang, Z. DeVito, M. Rai- son, A. Tejani, S. Chilamkurthy, B. Steiner, L. Fang, J. Bai, and S. Chintala. PyTorch: An Imperative Style, High-Performance Deep Learning Library. arXiv e-prints, page arXiv:1912.01703, Dec 2019.","title":"PyTorch: An Imperative Style, High-Performance Deep Learning Library","authors":"A Paszke, S Gross, F Massa, A Lerer, J Bradbury, G Chanan, T Killeen, Z Lin, N Gimelshein, L Antiga, A Desmaison, A Köpf, E Yang, Z Devito, M Raison, A Tejani, S Chilamkurthy, B Steiner, L Fang, J Bai, S Chintala","year":"2019","venue":null,"doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Russakovsky et al., 2015] O. Russakovsky, J. Deng, H. Su, J. Krause, S. Satheesh, S. Ma, Z. Huang, A. Karpathy, A. Khosla, M. Bernstein, A. Berg, and L. Fei-Fei. ImageNet Large Scale Vi- sual Recognition Challenge. Int. J. Comput. Vision, 115(3):211- 252, 2015.","title":"ImageNet Large Scale Visual Recognition Challenge","authors":"O Russakovsky, J Deng, H Su, J Krause, S Satheesh, S Ma, Z Huang, A Karpathy, A Khosla, M Bernstein, A Berg, L Fei-Fei","year":"2015","venue":"Int. J. Comput. Vision","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Shen et al., 2015] F. Shen, C. Shen, W. Liu, and H. T. Shen. Su- pervised discrete hashing. In CVPR, 2015.","title":"Supervised discrete hashing","authors":"F Shen, C Shen, W Liu, H T Shen","year":"2015","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Shen et al., 2019] Y. Shen, J. Qin, J. Chen, L. Liu, and F. Zhu. Embarrassingly Simple Binary Representation Learning. arXiv e-prints, page arXiv:1908.09573, Aug 2019.","title":"Embarrassingly Simple Binary Representation Learning","authors":"Y Shen, J Qin, J Chen, L Liu, F Zhu","year":"2019","venue":null,"doi":null,"url":null,"internal_url":null},{"position":29,"raw":"[Su et al., 2018] S. Su, C. Zhang, K. Han, and Y. Tian. Greedy hash: Towards fast optimization for accurate hash coding in CNN. In NeurIPS, pages 806-815, 2018.","title":"Greedy hash: Towards fast optimization for accurate hash coding in CNN","authors":"S Su, C Zhang, K Han, Y Tian","year":"2018","venue":"NeurIPS","doi":null,"url":null,"internal_url":null},{"position":30,"raw":"[Tang, 2013] Y. Tang. Deep Learning using Linear Support Vector Machines. arXiv e-prints, page arXiv:1306.0239, Jun 2013.","title":"Deep Learning using Linear Support Vector Machines","authors":"Y Tang","year":"2013","venue":null,"doi":null,"url":null,"internal_url":null},{"position":31,"raw":"[Wang et al., 2013] J. Wang, W. Liu, A. X. Sun, and Y. Jiang. Learning hash codes with listwise supervision. In ICCV, pages 3032-3039, 2013.","title":"Learning hash codes with listwise supervision","authors":"J Wang, W Liu, A X Sun, Y Jiang","year":"2013","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":32,"raw":"[Wang et al., 2015] Q. Wang, Z. Zhang, and S. Luo. Ranking pre- serving hashing for fast similarity search. In IJCAI, pages 3911- 3917, 2015.","title":"Ranking preserving hashing for fast similarity search","authors":"Q Wang, Z Zhang, S Luo","year":"2015","venue":"IJCAI","doi":null,"url":null,"internal_url":"/proceedings/2015/549"},{"position":33,"raw":"[Wang et al., 2016] X. Wang, Y. Shi, and K. Kitani. Deep super- vised hashing with triplet labels. In ACCV, pages 70-84, 2016.","title":"Deep supervised hashing with triplet labels","authors":"Wang","year":"2016","venue":"ACCV","doi":null,"url":null,"internal_url":null},{"position":34,"raw":"[Wang et al., 2018] J. Wang, T. Zhang, J. Song, N. Sebe, and H. T. Shen. A survey on learning to hash. T-PAMI, 40(4):769-790, 2018.","title":"A survey on learning to hash","authors":"J Wang, T Zhang, J Song, N Sebe, H T Shen","year":"2018","venue":"T-PAMI","doi":null,"url":null,"internal_url":null},{"position":35,"raw":"[Yang et al., 2018] H. Yang, K. Lin, and C. Chen. Supervised learn- ing of semantics-preserving hash via deep convolutional neural networks. T-PAMI, 40(2):437-451, 2018.","title":"Supervised learning of semantics-preserving hash via deep convolutional neural networks","authors":"H Yang, K Lin, C Chen","year":"2018","venue":"T-PAMI","doi":null,"url":null,"internal_url":null},{"position":36,"raw":"[Yao et al., 2016] T. Yao, F. Long, T. Mei, and Y. Rui. Deep semantic-preserving and ranking-based hashing for image re- trieval. In IJCAI, 2016.","title":"Deep semantic-preserving and ranking-based hashing for image retrieval","authors":"T Yao, F Long, T Mei, Y Rui","year":"2016","venue":"IJCAI","doi":null,"url":null,"internal_url":"/proceedings/2016/553"},{"position":37,"raw":"[Zhao et al., 2015] F. Zhao, Y. Huang, L. Wang, and T. Tan. Deep semantic ranking based hashing for multi-label image retrieval. In CVPR, 2015.","title":"Deep semantic ranking based hashing for multi-label image retrieval","authors":"F Zhao, Y Huang, L Wang, T Tan","year":"2015","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":38,"raw":"[Zhuang et al., 2016] B. Zhuang, G. Lin, C. Shen, and I. Reid. Fast training of triplet-based deep binary embedding networks. In CVPR, 2016.","title":"Fast training of triplet-based deep binary embedding networks","authors":"B Zhuang, G Lin, C Shen, I Reid","year":"2016","venue":"CVPR","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}