基于上下文约束的人脸聚类算法

摘要/Abstract

摘要： 人脸聚类将属于同一身份的人脸图片自动分到同一组,可用于人脸标注和图像管理等领域。传统的聚类算法的准确率很高,其召回率往往较低。为了解决这个问题,文中提出了一种加入三角约束关系和上下文约束的聚类算法。该聚类算法基于条件随机场模型,结合三角约束关系,同时考虑图像中常见的上下文约束,分别在聚类迭代过程和初步聚类之后添加最大相似度约束和共同出现约束,并对簇类进行整合。实验结果显示,结合上下文约束的条件随机场模型可以快速进行人脸聚类,同时还能保证在较高准确率的基础上有较高的召回率,从而提高整体的聚类效果。

关键词: 人脸聚类, 三角约束关系, 上下文约束, 条件随机场

Abstract: Face clustering which aims to automatically divide face images of the same identity into the same cluster,can be applied in a wide range of applications such as face annotation,image management,etc.The traditional face clustering algorithms can achieve good precision,but low recall.To handle this issue,this paper proposed a novel clustering algorithm with triangular constraints and context constraints.The proposed algorithm based on conditional random field model takes triangular constraints as well as common context constraints into accountin images.During the clustering iteration and after preliminary clustering,maximum similarity and people co-occurrence constraints are considered to merge the initial clusters.Experimental results reveal that the proposed face clustering algorithm can group faces efficiently,and improve recall with the high precision,and accordingly enhance the overall clustering performance.

Key words: Conditional random field, Context constraints, Face clustering, Triangular constraints

中图分类号:

TP391.4

罗恒利, 王文博, 葛宏孔. 基于上下文约束的人脸聚类算法[J]. 计算机科学, 2019, 46(11A): 260-263. https://doi.org/

LUO Heng-li, WANG Wen-bo, GE Hong-kong. Face Clustering Algorithm Based on Context Constraints[J]. Computer Science, 2019, 46(11A): 260-263. https://doi.org/

参考文献

[1]VRETOS N,SOLACHIDIS V,PITAS I.A mutual information based face clustering algorithm for movie content analysis[J].Image & Vision Computing,2011,29(10):693-705.
[2]ZHANG Y,TANG Z,WU B,et al.A Coupled Hidden Conditional Random Field Model for Simultaneous Face Clustering and Naming in Videos[J].IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society,2016,25(12):5780-5792.
[3]SCHROFF F,KALENICHENKO D,PHILBIN J.Facenet:A unified embedding for face recognition and clustering[C]∥Proceedings of the IEEE Conference on Computervision and Pattern Recognition.2015:815-823.
[4]ZHANG Z P,LUO P,CHEN C L,et al.Joint face representation adaptation and clustering invideos[C]∥European Conference on Computer Vision.Springer,2016:236-251.
[5]SUN Y,CHEN Y H,WANG X G,et al.Deep learning face representation by joint identificationverification[C]∥Advances in Neural Information Processing Systems.2014:1988-1996.
[6]TURK M,PENTLAND A.Eigenfaces for recognition[J].JCogn Neurosci,1991,3(1):71-86.
[7]AHONEN T,HADID A,PIETIKAINEN M.Face Description with Local Binary Patterns:Application to Face Recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(12):2037-2041.
[8]TAN H,YANG B,MA Z.Face recognition based on the fusion of global and local HOG features of face images[J].IET Computer Vision,2014,8(3):224-234.
[9]SUN Y,WMG X,TANG X.Deep learning face representationfrom predicting10,000 classes[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2014:1891-1898.
[10]http://cmusatyalab.github.io/openface.
[11]MACQUEEN J.Some Methods for Classification and Analysis ofMultiVariate Observations[C]∥Proc of Berkeley Symposium on Mathematical Statistics & Probability.1965.
[12]FREY B J,DUECK D.Clustering by Passing Messages Between Data Points[J].Science,2007,315(5814):972-976.
[13]SHI J,MALIK J.Normalized cuts and image segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(8):888-905.
[14]NG A Y,JORDAN M I,WEISS Y.Y:On spectral clustering:analysis and an algorithm[J].Proc Nips,2001,14:849-856.
[15]RODRIGUEZ A,LAIO A.Mechine learning,Clustering by fast search and find of density peaks[J].Science,2014,344(6191):1492.
[16]SHI Y,OTTO C,JAIN A K.Face Clustering:Representation and Pairwise Constraints[J].IEEE Transactions on Information Forensics & Security,2017,PP(99):1-1.
[17]GALLAGHER A C,CHEN T.Clothing cosegmentation for recognizing people[C]∥IEEE Conference on Computer Vision and Pattern Recognition,2008(CVPR 2008).IEEE,2008.
[18]https://github.com/seetaface/SeetaFaceEngine.
[19]AMIGÓE,GONZALO J,ARTILES J,et al.A comparison of extrinsic clustering evaluation metrics based on formal constraints[J].Information Retrieval,2009,12(4):461-486.
[20]ZHANG L,KALASHNIKOV D V,MEHROTRA S.A unified framework for context assisted face clustering[C]∥Acm Conference on International Conference on Multimedia Retrieval.2013.

相关文章 15

[1]	丁锋, 孙晓. 基于注意力机制和BiLSTM-CRF的消极情绪意见目标抽取 Negative-emotion Opinion Target Extraction Based on Attention and BiLSTM-CRF 计算机科学, 2022, 49(2): 223-230. https://doi.org/10.11896/jsjkx.210100046
[2]	王文博, 罗恒利. 基于图卷积神经网络的完全图人脸聚类 Complete Graph Face Clustering Based on Graph Convolution Network 计算机科学, 2021, 48(11A): 275-277. https://doi.org/10.11896/jsjkx.201200102
[3]	周鹏程,龚声蓉,钟珊,包宗铭,戴兴华. 基于深度特征融合的图像语义分割 Image Semantic Segmentation Based on Deep Feature Fusion 计算机科学, 2020, 47(2): 126-134. https://doi.org/10.11896/jsjkx.190100119
[4]	孙国梓, 吕建伟, 李华康. 基于编辑距离的多实体可信确认算法 MeTCa:Multi-entity Trusted Confirmation Algorithm Based on Edit Distance 计算机科学, 2020, 47(12): 327-331. https://doi.org/10.11896/jsjkx.191100176
[5]	林增敏, 洪朝群, 庄蔚蔚. 融合人脸跟踪和聚类的人脸图像去重方法 Face Image Deduplication Based on Fusion of Face Tracking and Clustering 计算机科学, 2020, 47(11A): 615-619. https://doi.org/10.11896/jsjkx.200400142
[6]	王子牛, 姜猛, 高建瓴, 陈娅先. 基于BERT的中文命名实体识别方法 Chinese Named Entity Recognition Method Based on BERT 计算机科学, 2019, 46(11A): 138-142.
[7]	陈伟, 吴友政, 陈文亮, 张民. 基于BiLSTM-CRF的关键词自动抽取 Automatic Keyword Extraction Based on BiLSTM-CRF 计算机科学, 2018, 45(6A): 91-96.
[8]	杨旭华,彭朋. 基于条件随机场和低采样率浮动车数据的地图匹配算法 Map Matching Algorithm Based on Conditional Random Fields and Low-sampling-rate Floating Car Data 计算机科学, 2016, 43(Z6): 68-72. https://doi.org/10.11896/j.issn.1002-137X.2016.6A.015
[9]	赵世瑜,线岩团,郭剑毅,余正涛,洪玄贵,王红斌. 基于条件随机场的泰语音节切分方法 Thai Syllable Segmentation Based on Conditional Random Fields 计算机科学, 2016, 43(3): 54-56. https://doi.org/10.11896/j.issn.1002-137X.2016.03.010
[10]	孙晓,孙重远,任福继. 基于深层结构模型的新词发现与情感倾向判定 New Word Detection and Emotional Tendency Judgment Based on Deep Structured Model 计算机科学, 2015, 42(9): 208-213. https://doi.org/10.11896/j.issn.1002-137X.2015.09.040
[11]	马萌,唐卓,李仁发,熊燎特. 基于条件随机场的改进型BLP访问控制模型 Improved BLP Model Based on CRFs 计算机科学, 2015, 42(8): 138-144.
[12]	冯蕴天张宏军郝文宁. 面向军事文本的命名实体识别 Named Entity Recognition for Military Text 计算机科学, 2015, 42(7): 15-18. https://doi.org/10.11896/j.issn.1002-137X.2015.07.004
[13]	刘建伟,黎海恩,罗雄麟. 概率图模型表示理论 Representation Theory of Probabilistic Graphical Models 计算机科学, 2014, 41(9): 1-17. https://doi.org/10.11896/j.issn.1002-137X.2014.09.001
[14]	邱泉清,苗夺谦,张志飞. 中文微博命名实体识别 Named Entity Recognition on Chinese Microblog 计算机科学, 2013, 40(6): 196-198.
[15]	赵海勇，贾保先. 基于轮廓特征的人体行为识别 Human Action Recognition Using Image Contour 计算机科学, 2013, 40(2): 312-314.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed