计算机科学

计算机科学 ›› 2023, Vol. 50 ›› Issue (9): 192-201.doi: 10.11896/jsjkx.220900133

• 数据库&大数据&数据科学 • 上一篇    下一篇

基于对比学习的超多类深度图像聚类模型

胡深1,3, 钱宇华1,2,3, 王婕婷1,3, 李飞江1,3, 吕维1,3   

  1. 1 山西大学计算机与信息技术学院 太原 030006
    2 山西大学计算智能与中文信息处理教育部重点实验室 太原 030006
    3 山西大学大数据科学与产业研究院 太原 030006
  • 收稿日期:2022-08-02 修回日期:2022-10-10 出版日期:2023-09-15 发布日期:2023-09-01
  • 通讯作者: 钱宇华(jinchengqyh@126.com)
  • 作者简介:(hushenshen@aliyun.com)
  • 基金资助:
    国家自然科学基金重点项目(62136005);国家重点研发计划(2021ZD0112400);国家自然科学基金青年科学基金(62106132);山西省三晋学者项目资助;山西省基础研究计划(20210302124271,202103021223026)

Super Multi-class Deep Image Clustering Model Based on Contrastive Learning

HU Shen1,3, QIAN Yuhua1,2,3, WANG Jieting1,3, LI Feijiang1,3, LYU Wei1,3   

  1. 1 School of Computer and Information Technology,Shanxi University,Taiyuan 030006,China
    2 Shanxi University Key Laboratory of Computational Intelligence and Chinese Information Processing,Ministry of Education,Taiyuan 030006,China
    3 Institute of Big Data Science and Industry,Shanxi University,Taiyuan 030006,China
  • Received:2022-08-02 Revised:2022-10-10 Online:2023-09-15 Published:2023-09-01
  • About author:HU Shen,born in 1997,postgraduate,is a student member of China Computer Federation.His main research interests include self-supervised learning and super multi-class image clustering.
    QIAN Yuhua,born in 1976,Ph.D,professor,is a member of China Computer Federation.His main research interests include artificial intelligence,big data,machine learning and data mining.
  • Supported by:
    Key Program of the National Natural Science Foundation of China(62136005), National Key Research and Development Program of China(2021ZD0112400),Young Scientists Fund of the National Natural Science Foundation of China(62106132),Program for the San Jin Young Scholars of Shanxi and Shanxi Provincial Research Foundation for Basic Research,China(20210302124271,202103021223026).

PDF (PC)

1303

摘要： 图像聚类通过表征学习对图像数据降维并提取有效特征而后进行聚类分析。当图像数据存在超多类别时,数据分布的复杂性和类簇的密集性严重影响了现有方法的实用性。为此,提出了基于对比学习的超多类深度图像聚类模型,主要分为3个阶段:首先,改进对比学习方法训练特征模型以使类簇分布均匀;其次,基于语义相似性原则多视角挖掘实例语义最近邻信息;最后,将实例及其最近邻作为自监督信息训练聚类模型。根据实验类型的不同,设计了消融实验和对比实验。在消融实验中,证明了所提方法使类簇均匀分布在映射空间,并可靠挖掘语义最近邻信息。在对比实验中,将其与先进算法在7个基准数据集上进行了比较,在ImageNet-200类数据集上,其准确率比目前先进方法提升了10.6%;在ImageNet-1000类数据集上,其准确率比目前先进算法提升了9.2%。

关键词: 超多类聚类, 对比学习, 特征模型, 语义相似性, 图像聚类

Abstract: Image clustering reduces the dimensionality of image data,extracts effective features through representation learning,and performs cluster analysis.When there are many categories of image data,the complexity of data distribution and the density of clusters seriously affect the practicability of existing methods.To this end,this paper proposes a super-multi-class deep image clustering model based on contrastive learning,which is mainly divided into three stages:firstly,improving the contrastive lear-ning method to train the feature model to make the cluster distribution uniform;secondly,based on the principle of semantic similarity,the perspective mines instance semantic nearest neighbor information;and finally,the instance and its nearest neighbors are used as self-supervised information to train a clustering model.According to the different types of experiments,ablation experiments and contrast experiments are designed in this paper.The ablation experiments prove that the proposed method could make the clusters evenly distributed in the mapping space and mine the semantic nearest neighbor information reliably.In the comparative experiments,it's compared with the advanced algorithms on 7 benchmark datasets.On the ImageNet-200 class dataset,it's accuracy is 10.6% higher than the advanced method.It's accuracy rate on the ImageNet-1000 class dataset is higher than that of the advanced algorithm,which improves by 9.2%.

Key words: Super multi-class clustering, Contrastive learning, Feature model, Semantic similarity, Image clustering

中图分类号: 

  • TP391
[1]REN X T,ZHAO J J,QIANG Y,et al.Lung Cancer SubtypeRecognition with Unsupervised Learning Combining Paired Learning and Image Clustering[J].Computer Science,2020,47(10):200-206.
[2]ZHU Z Q,GENG H J,QIAN Y H.Line-Segment Clustering Algorithm for Chemical Structure[J].Computer Science,2022,49(5):113-119.
[3]LU K,DING Z,GE S.Sparse-Representation-Based Graph Embedding for Traffic Sign Recognition[J].IEEE Transactions on Intelligent Transportation Systems,2012,13(4):1515-1524.
[4]LIU X,WANG G Y,LUO X B.Multi-granularity Clustering of Remote Sensing Image Based on Gaussian Cloud Transformation[J].Computer Science,2017,44(9):23-27,52.
[5]HUANG M,LI T,WEN X,et al.Induction Motor ParameterIdentification Incorporating Image Recognition and Cluster Ana-lysis[J].Journal of Chongqing University of Technology(Natural Science),2022,36(9):195-201.
[6]PARK S,HAN S,KIM S,et al.Improving Unsupervised Image Clustering With Robust Learning [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.IEEE Press,2021:15750-15758.
[7]NIU C,SHAN H,WANG G.SPICE:Semantic Pseudo-labeling for Image Clustering[J].arXiv:2103.09382,2021.
[8]COATES A,LEE H,NG A Y.An Analysis of Single-LayerNetworks in Unsupervised Feature Learning[C]//Proceedings of the fourteenth International Conference on Artificial Intelligence and Statistics.Lauderdale:JMLR Press,2011:215-223.
[9]KRIZHEVSKY A.Learning Multiple Layers of Features fromTiny Images[R].University of Toronto,Technical Report, 2009.
[10]TAO Y,TAKAGI K,NAKATA K.Clustering-friendly Representation Learning via Instance Discrimination and Feature Decorrelation[C]//Proceedings of the International Conference on Learning Representations.ICLR Press,2021.
[11]HUANG Z,CHEN J,ZHANG J,et al.Learning Representation for Clustering via Prototype Scattering and Positive Sampling[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2023,45(6):7509-7524.
[12]DESHMUKH A A,REGATTI J R,MANAVOGLU E,et al.Representation Learning for Clustering via Building Consensus[J].Springer Machine Learning Journal,2022,111(12):4601-4638.
[13]CHANG J,WANG L,MENG G,et al.Deep Adaptive ImageClustering[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.Venice:IEEE Press,2017:5880-5888.
[14]DENG J,DONG W,SOCHER R,et al.ImageNet:A Large-Scale Hierarchical Image Database[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Miami:IEEE Press,2009:248-255.
[15]LI F,QIAN Y,WANG J,et al.Clustering method based on sample's stability[J].SCIENTIA SINICA Informationis,2020,50(8):1239-1254.
[16]WANG J,QIAN Y,LI F,et al.Support Vector Machine withEliminating the Random Consistency[J].Journal of Computer Research and Development,2020,57(8):1581-1593.
[17]LI F,QIAN Y,WANG J,et al.Clustering ensemble based on sample's stability[J].Artificial Intelligence,2019,273:37-55.
[18]WANG J,QIAN Y,LI F,et al.Generalization Performance of Pure Accuracy and Its Application in Selective Ensemble Lear-ning[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2023,45(2):1798-1816.
[19]ZHANG H,ZHANG J,HUANG J.Multi--label Image Classification Model Based on Graph Attention Network[J].Journal of Chongqing Technology and Business University(Natural Science Edition),2022,39(1):34-41.
[20]MENG L,TAN A H,MIAO C.Salience-aware adaptive resonance theory for large-scale sparse data clustering[J].Neural Networks,2019,120:143-157.
[21]MENG L,TAN A H,WUNSCH D C.Vigilance adaptation inadaptive resonance theory[C]//Proceedings of the the 2013 International Joint Conference on Neural Networks.IEEE Press,2013:1-7.
[22]MENG L,TAN A H,XU D.Semi-Supervised HeterogeneousFusion for Multimedia Data Co-Clustering[J].IEEE Transactions on Knowledge and Data Engineering,2014,26(9):2293-2306.
[23]MENG L,TAN A H,WUNSCH D C.Adaptive Scaling of Cluster Boundaries for Large-Scale Social Media Data Clustering[J].IEEE Transactions on Neural Networks and Learning Systems,2016,27(12):2656-2669.
[24]CARPENTER G A,GROSSBERG S.A Massively Parallel Ar-chitecture for a Self-Organizing Neural Pattern Recognition Machine[J].Computer Vision,Graphics,and Image Processing,1987,37(1):54-115.
[25]CARPENTER G A,GROSSBERG S,ROSEN D B.Fuzzy ART:Fast stable learning and categorization of analog patterns by an adaptive resonance system[J].Neural Networks,1991,4(6):759-771.
[26]HUANG P,HUANG Y,WANG W,et al.Deep Embedding Network for Clustering[C]//Proceedings of the 22nd International Conference on Pattern Recognition.Washington:IEEE Press,2014:1532-1537.
[27]XIE J,GIRSHICK R,FARHADI A.Unsupervised Deep Em-bedding for Clustering Analysis[C]//Proceedings of the International Conference On Machine Learning.New York:ACM Press,2016:478-487.
[28]YANG B,FU X,SIDIROPOULOS N D,et al.TowardsK-means-friendly Spaces:Simultaneous Deep Learning and Clustering[C]//Proceedings of the International Conference On Machine Learning.New York:ACM Press,2017:3861-3870.
[29]JI P,ZHANG T,LI H,et al.Deep Subspace Clustering Networks[C]//Proceedings of the Annual Conference on Neural Information Processing Systems.MIT Press,2017.
[30]ZHOU P,HOU Y,FENG J.Deep Adversarial Subspace Clustering[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Salt Lake City:IEEE Press,2018:1596-1604.
[31]DIZAJI K G,HERANDI A,DENG C,et al.Deep Clustering via Joint Convolutional Autoencoder Embedding and Relative Entropy Minimization[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.Venice:IEEE Press,2017:5747-5756.
[32]YANG J,PARIKH D,BATRA D.Joint Unsupervised Learning of Deep Representations and Image Clusters[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE Press,2016:5147-5156.
[33]CARON M,BOJANOWSKI P,JOULIN A,et al.Deep Clustering for Unsupervised Learning of Visual Features[C]//Proceedings of the European conference on computer vision.Cham:Springer International Publishing,2018:139-156.
[34]WU J,LONG K,WANG F,et al.Deep Comprehensive Correlation Mining for Image Clustering[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.Seoul:IEEE Press,2019:8149-8158.
[35]JI X,VEDALDI A,HENRIQUES J.Invariant Information Clustering for Unsupervised Image Classification and Segmentation[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.Seoul:IEEE Press,2019:9864-9873.
[36]HUANG J,GONG S,ZHU X.Deep Semantic Clustering byPartition Confidence Maximisation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Seattle:IEEE Press,2020:8846-8855.
[37]VAN GANSBEKE W,VANDENHENDE S,GEORGOULIS S,et al.SCAN:Learning to Classify Images Without Labels[C]//Proceedings of the European conference on computer vision.Cham:Springer International Publishing,2020:268-285.
[38]LI Y,HU P,LIU Z,et al.Contrastive Clustering[C]//Procee-dings of the AAAI Conference on Artificial Intelligence.Menlo Park:AAAI,2021:8547-8555.
[39]HE K,FAN H,WU Y,et al.Momentum Contrast for Unsupervised Visual Representation Learning[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Seattle:IEEE Press,2020:9729-9738.
[40]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.New York:ACM Press,2020:1597-1607.
[41]GRILL J B,STRUB F,ALTCHÉ F,et al.Bootstrap your own latent:A new approach to self-supervised Learning[C]//Proceedings of the Annual Conference on Neural Information Processing Systems.MIT Press,2020:21271-21284.
[42]CHEN X,HE K.Exploring Simple Siamese RepresentationLearning[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.IEEE Press,2021:15750-15758.
[43]WANG T,ISOLA P.Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere[C]//Proceedings of the International Conference On Machine Learning.New York:ACM Press,2020:9929-9939.
[44]DEVRIES T,TAYLOR G W.Improved Regularization of Con-volutional Neural Networks with Cutout[J].arXiv:1708.04552,2017.
[45]CUBUK E D,ZOPH B,SHLENS J,et al.Randaugment:Practical automated data augmentation with a reduced search space[C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops.IEEE Press,2020:3008-3017.
[46]WANG F,LIU H.Understanding the Behaviour of Contrastive Loss[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Nashville:IEEE Press,2021:2495-2504.
[47]SOHN K,BERTHELOT D,LI C L,et al.FixMatch:Simplifying Semi-Supervised Learning with Consistency and Confıdence[J].Advances in neural information processing systems,2020,33:596-608.
[48]WANG Y,HUANG H,RUDIN C,et al.Understanding HowDimension Reduction Tools Work:An Empirical Approach to Deciphering t-SNE,UMAP,TriMAP,and PaCMAP for Data Visualization[J].Journal of Machine Learning Research,2021,22(201):1-73.
[49]ZELNIK-MANOR L,PERONA P.Self-Tuning Spectral Clustering[C]//Proceedings of the Annual Conference on Neural Information Processing Systems.MIT Press,2004.
[50]BENGIO Y,LAMBLIN P,POPOVICI D,et al.Greedy Layer-Wise Training of Deep Networks[C]//Proceedings of the Annual Conference on Neural Information Processing Systems.MIT Press,2006.
[51]KINGMA D P,WELLING M.Auto-Encoding Variational Bayes[C]//Proceedings of the International Conference on Learning Representations.ICLR Press,2014.
[52]ZHONG H,WU J,CHEN C,et al.Graph Contrastive Clustering[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.Montreal:IEEE Press,2021:9204-9213.
[53]LI J,ZHOU P,XIONG C,et al.Prototypical Contrastive Lear-ning of Unsupervised Representations[C]//Proceedings of the International Conference on Learning Representations.ICLR Press,2021.
[54]TSAI T W,LI C,ZHU J.MiCE:Mixture of Contrastive Experts for Unsupervised Image Clustering[C]//Proceedings of the International Conference on Learning Representations.ICLR Press,2021.
[55]COHEN N,HOSHEN Y.The Single-Noun Prior for ImageClustering[J].arXiv:2104.03952,2021.
[56]CHEN X,FAN H,GIRSHICK R,et al.Improved Baselines with Momentum Contrastive Learning[J].arXiv:2003.04297,2020.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发