计算机科学 ›› 2022, Vol. 49 ›› Issue (10): 159-168.doi: 10.11896/jsjkx.210800050
苗壮, 王亚鹏, 李阳, 王家宝, 张睿, 赵昕昕
MIAO Zhuang, WANG Ya-peng, LI Yang, WANG Jia-bao, ZHANG Rui, ZHAO Xin-xin
摘要: 为了提高无监督哈希学习的性能,实现鲁棒的哈希图像检索,提出了一种鲁棒的双教师自监督蒸馏哈希学习方法。该方法包括自监督双教师学习和鲁棒哈希学习两个阶段:第一阶段设计了一种改进的聚类算法,有效提高了硬伪标签的标注精度,而后通过微调教师网络得到了图像的初始软伪标签;第二阶段提出了一种结合混合去噪和双教师共识去噪策略的软伪标签去噪方法,有效去除了初始软伪标签中的噪声,而后利用蒸馏学习将双教师网络中的信息通过去噪软伪标签传递给学生网络,进而获得无标签图像的鲁棒哈希码。在CIFAR-10,FLICKR25K和EuroSAT上进行了实验,实验结果表明,与TBH方法相比,在CIFAR-10上所提方法的MAP平均提高了18.6%;与DistillHash方法相比,在FLICKR25K上所提方法的MAP平均提高了2.4%;与ETE-GAN方法相比,在EuroSAT上所提方法的MAP平均提高了18.5%。
中图分类号:
[1]WANG J,KUMAR S,CHANG S F.Semi-supervised hashing for large-scale search[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2012,34(12):2393-2406. [2]SHARIF U,MEHMOOD Z,MAHMOOD T,et al.Scene analysis and search using local features and support vector machine for effective content-based image retrieval[J].Artificial Intelligence Review,2019,52(2):901-925. [3]PENG Y,ZHANG J,YE Z.Deep reinforcement learning forimage hashing[J].IEEE Transactions on Multimedia,2019,22(8):2061-2073. [4]ZHANG B,QIAN J.Autoencoder-based unsupervised clustering and hashing[J].Applied Intelligence,2021,51(1):493-505. [5]GOODFELLOW I J,POUGET-ABADIE J,MIRZA M,et al.Generative adversarial networks[C]//Proceedings of Advances in Neural Information Processing Systems.2014:2672-2680. [6]SHEN Y,QIN J,CHEN J,et al.Auto-encoding twin-bottleneck hashing[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2020:2815-2824. [7]CHEN X,LU C.An end-to-end adversarial hashing method for unsupervised multispectral remote sensing image retrieval[C]//Proceedings of the IEEE International Conference on Image Processing.IEEE,2020:1536-1540. [8]SONG J,HE T,GAO L,et al.Binary generative adversarial net-works for image retrieval[C]//Proceedings of the AAAI Conference on Artificial Intelligence.AAAI,2018:394-401. [9]YANG E,LIU T,DENG C,et al.Distillhash:unsupervised deep hashing by distilling data pairs[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2019:2946-2955. [10]YANG E,DENG C,LIU T,et al.Semantic structure-based unsupervised deep hashing[C]//Proceedings of the International Joint Conference on Artificial Intelligence.2018:1064-1070. [11]DIZAJI G,ZHENG F,SADOUGHI N,et al.Unsupervised deep generative adversarial hashing network[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2018:3664-3673. [12]LIN K,LU J,CHEN C S,et al.Learning compact binary de-scriptors with unsupervised deep neural networks[C]//Procee-dings of the IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2016:1183-1192. [13]DAI B,GUO R,KUMAR S,et al.Stochastic generative hashing[C]//Proceedings of the International Conference on Machine Learning.PMLR,2017:913-922. [14]LI Y,WANG Y,MIAO Z,et al.Contrastive self-supervised hashing with dual pseudo agreement[J].IEEE Access,2020,8:165034-165043. [15]HUANG J,DONG Q,GONG S,et al.Unsupervised deep lear-ning by neighbourhood discovery[C]//Proceedings of the International Conference on Machine Learning.PMLR,2019,97:2849-2858. [16]CHEN T,KORNBLITH S,NOROUZI M,et al.A simpleframework for contrastive learning of visual representations[C]//Proceedings of the International Conference on Machine Learning.PMLR,2020,119:1597-1607. [17]HE K,FAN H,WU Y,et al.Momentum contrast for unsupervised visual representation learning[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2020:9726-9735. [18]BACHMAN P,HJELM R,BUCHWALTER W.Learning representations by maximizing mutual information across views[C]//Proceedings of Advances in Neural Information Proces-sing Systems.2019:15509-15519. [19]CARON M,BOJANOWSKI P,JOULIN A,et al.Deep clustering for unsupervised learning of visual features[C]//Procee-dings of the European Conference on Computer Vision.Berlin:Springer,2018:139-156. [20]ASANO Y,RUPPRECHT C,VEDALDI A.Self-labelling viasimultaneous clustering and representation learning[C]//Proceedings of the International Conference on Learning Representations,OpenReview,2020. [21]CARON M,MISRA I,MAIRAL J,et al.Unsupervised learning of visual features by contrasting cluster assignments[C]//Proceedings of Advances in Neural Information Processing Systems.2020. [22]HAN B,YAO Q,LIU T,et al.A survey of label-noise representation learning:past,present and future[EB/OL].(2021-02-20) [2021-04-27].https://arxiv.org/abs/2011.04406. [23]REED S,LEE H,ANGUELOV D,et al.Training deep neuralnetworks on noisy labels with bootstrapping[C]//Proceedings of the International Conference on Learning Representations.2015. [24]SHU J,XIE Q,YI L,et al.Meta-weight-net:Learning an expli-cit mapping for sample weighting[C]//Proceedings of Advances in Neural Information Processing Systems.2019:1919-1930. [25]LI Y,YANG J,SONG Y,et al.Learning from noisy labels with distillation[C]//Proceedings of the International Conference on Computer Vision.IEEE,2017:1928-1936. [26]HUANG Z,ZOU Y,BHAGAVATULA V,et al.Comprehen-sive attention self-distillation for weakly-supervised object detection[C]//Proceedings of Advances in Neural Information Processing Systems.2020. [27]HINTON G,VINYALS O,DEAN J.Distilling the knowledge in a neural network[J].Computer Science,2015,14(7):38-39. [28]NORTHCUTT C,JIANG L,CHUANG I.Confident Learning:Estimating Uncertainty in Dataset Labels[J].Journal of Artificial Intelligence Research,2021,70:1373-1411. [29]WEISS Y,TORRALBA A,FERGUS R.Spectral hashing[C]//Proceedings of Advances in Neural Information Processing Systems,Curran Associates,Inc..2008:1753-1760. [30]ANDONI A,INDYK P.Near-optimal hashing algorithms for approximate nearest neighbor in high dimensions[J].Communications of the ACM,2008,51(1):117-122. [31]WANG X,ZHANG L,JING F,et al.AnnoSearch:Image auto-annotation by search[C]//Proceedings of the 2006 IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2006:1483-1490. [32]HEO J,LEE Y,HE J,et al.Spherical hashing[C]//Proceedings of the 2012 IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2012:2957-2964. [33]JIN Z,LI C,LIN Y,et al.Density sensitive hashing[J].IEEE Transactions on Cybernetics,2014,44(8):1362-1371. [34]GONG Y,LAZEBNIK S,GORDO A,et al.Iterative quantiza-tion:A procrustean approach to learning binary codes for large-scale image retrieval[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2013,35(12):2916-2929. [35]WEI Y,TIAN D,SHI J,et al.Optimizing non-differentiablemetrics for hashing[J].IEEE Access,2021,9:14351-14357. [36]RAGINSKY M,LAZEBNIK S.Locality-sensitive binary codesfrom shift-invariant kernels[C]//Proceedings of Advances in Neural Information Processing Systems,Curran Associates,Inc..2009:1509-1517. [37]XIA Y,HE K,KOHLI P,et al.Sparse projections for high-dimensional binary codes[C]//Proceedings of the IEEE Confe-rence on Computer Vision and Pattern Recognition.IEEE,2015:3332-3339. |
[1] | 曹晓雯, 梁美玉, 鲁康康. 基于细粒度语义推理的跨媒体双路对抗哈希学习模型 Fine-grained Semantic Reasoning Based Cross-media Dual-way Adversarial Hashing Learning Model 计算机科学, 2022, 49(9): 123-131. https://doi.org/10.11896/jsjkx.220600011 |
[2] | 杜航原, 李铎, 王文剑. 一种面向电商网络的异常用户检测方法 Method for Abnormal Users Detection Oriented to E-commerce Network 计算机科学, 2022, 49(7): 170-178. https://doi.org/10.11896/jsjkx.210600092 |
[3] | 楚玉春, 龚航, 王学芳, 刘培顺. 基于YOLOv4的目标检测知识蒸馏算法研究 Study on Knowledge Distillation of Target Detection Algorithm Based on YOLOv4 计算机科学, 2022, 49(6A): 337-344. https://doi.org/10.11896/jsjkx.210600204 |
[4] | 程祥鸣, 邓春华. 基于无标签知识蒸馏的人脸识别模型的压缩算法 Compression Algorithm of Face Recognition Model Based on Unlabeled Knowledge Distillation 计算机科学, 2022, 49(6): 245-253. https://doi.org/10.11896/jsjkx.210400023 |
[5] | 黄仲浩, 杨兴耀, 于炯, 郭亮, 李想. 基于多阶段多生成对抗网络的互学习知识蒸馏方法 Mutual Learning Knowledge Distillation Based on Multi-stage Multi-generative Adversarial Network 计算机科学, 2022, 49(10): 169-175. https://doi.org/10.11896/jsjkx.210800250 |
[6] | 田嵩旺, 蔺素珍, 杨博. 基于多判别器的多波段图像自监督融合方法 Multi-band Image Self-supervised Fusion Method Based on Multi-discriminator 计算机科学, 2021, 48(8): 185-190. https://doi.org/10.11896/jsjkx.200600132 |
[7] | 王春静, 刘丽, 谭艳艳, 张化祥. 基于模糊颜色特征和模糊相似度的图像检索方法 Image Retrieval Method Based on Fuzzy Color Features and Fuzzy Smiliarity 计算机科学, 2021, 48(8): 191-199. https://doi.org/10.11896/jsjkx.200800202 |
[8] | 王晓飞, 周超, 刘利刚. 基于特征聚类的轻量级图像搜索系统 Lightweight Image Retrieval System Based on Feature Clustering 计算机科学, 2021, 48(2): 148-152. https://doi.org/10.11896/jsjkx.191200104 |
[9] | 俞亮, 魏永丰, 罗国亮, 邬昌兴. 基于知识蒸馏的隐式篇章关系识别 Knowledge Distillation Based Implicit Discourse Relation Recognition 计算机科学, 2021, 48(11): 319-326. https://doi.org/10.11896/jsjkx.201000099 |
[10] | 徐行, 孙嘉良, 汪政, 杨阳. 基于特征变换的图像检索对抗防御 Feature Transformation for Defending Adversarial Attack on Image Retrieval 计算机科学, 2021, 48(10): 258-265. https://doi.org/10.11896/jsjkx.200800222 |
[11] | 王润正, 高见, 黄淑华, 仝鑫. 基于知识蒸馏的恶意代码家族检测方法 Malicious Code Family Detection Method Based on Knowledge Distillation 计算机科学, 2021, 48(1): 280-286. https://doi.org/10.11896/jsjkx.200900099 |
[12] | 余孟池, 牟甲鹏, 蔡剑, 徐建. 噪声标签重标注方法 Noisy Label Classification Learning Based on Relabeling Method 计算机科学, 2020, 47(6): 79-84. https://doi.org/10.11896/jsjkx.190600041 |
[13] | 焦扬, 杨传颖, 石宝. 基于SVM相关反馈的鞋印图像检索算法 Relevance Feedback Method Based on SVM in Shoeprint Images Retrieval 计算机科学, 2020, 47(11A): 244-247. https://doi.org/10.11896/jsjkx.200400032 |
[14] | 孙伟, 赵玉普. 增强旋转不变LBP算法及其在图像检索中的应用 Enhanced Rotation Invariant LBP Algorithm and Its Application in Image Retrieval 计算机科学, 2019, 46(7): 263-267. https://doi.org/10.11896/j.issn.1002-137X.2019.07.040 |
[15] | 彭金喜, 苏远歧, 薛笑荣. 基于深度学习和同生矩阵的SAR图像纹理特征检索方法 SAR Image Feature Retrieval Method Based on Deep Learning and Synchronic Matrix 计算机科学, 2019, 46(6A): 196-199. |
|