域通用和域分离字典对学习的行人重识别算法

doi:10.3778/j.issn.1002-8331.2012-0276

摘要/Abstract

摘要： 为克服不同相机视角之间的域偏移问题，提出一种基于域通用和域分离字典对学习的跨视角行人重识别算法。具体地，基于来自同一相机视角下的行人共享相同的域，并且同一视角中每个行人图像所携带的域信息在短时间内具有一致性，将同一视角下的行人图像分解为特定视角的域信息分量和域分离的行人外观特征分量，提出一个判别字典学习模型以创建用于描述域信息分量的域通用字典和描述行人外观分量的域分离字典。由于来自同一相机视角下的图像具有域相似性，因此通过低秩正则化来细化用于表示域信息的字典。为了进一步提高学习字典的判别能力，在算法中约束相同视角、相同身份的多幅图像的编码系数具有很强的相似性。此外，采用一种新颖的扩展正则化方法来解决不同行人相似外貌特征和同一行人不同外貌特征的视觉外观歧义问题。在四个具有挑战性的数据集上进行实验，结果表明域通用和域分离字典对学习的算法相对于一些现有最新算法更具有效性和优越性。

关键词: 行人重识别, 域通用字典, 扩展正则化, 域信息

Abstract: In order to overcome the problem of domain-shift between different camera views, a domain-common and domain-separation dictionary pair learning approach for cross?view person re?identification is proposed. Specifically, based on the fact that pedestrians from the same camera view share the same domain, and that the information of domain carried by each pedestrian image in the same view has consistency for short time, the person image from the same view is decomposed into specific perspective domain information components and domain-separation pedestrian appearance feature components, and a discriminative dictionary learning model is developed to create a domain-common dictionary for characterizing the domain information components and a domain-separation dictionary for characterizing the pedestrian appearance components. Since images from the same camera view have domain similarity, dictionary used to represent domain information is refined by low-rank regularization. Besides, to further improve discrimination ability of learned dictionaries, certain constraint is proposed in the algorithm, that is, the coding coefficients of multiple images with the same view and identity to have strong similarity. Furthermore, a novel expansion regularization is used to solve the visual appearance ambiguity of similar appearance features of different pedestrians and different appearance features of the same pedestrian. Finally, experiments on four challenging datasets demonstrate the effectiveness and the superiority of domain-common and domain-separation dictionary pair learning approach against some state-of-the-art methods.

Key words: person re-identification, domain-common dictionary, expansion regularization, domain information

颜悦, 严双林, 颜昌沁. 域通用和域分离字典对学习的行人重识别算法[J]. 计算机工程与应用, 2022, 58(15): 141-152.

YAN Yue, YAN Shuanglin, YAN Changqin. Domain-Common and Domain-Separation Dictionary Pair Learning for Person Re-Identification[J]. Computer Engineering and Applications, 2022, 58(15): 141-152.

参考文献

[1] CHEN Y C，ZHENG W S，LAI J H，et al.An asymmetric distance model for cross-view feature mapping in person re-identification[J].IEEE Transactions on Circuits and Systems for Video Technology，2017，27（8）：1661-1675.
[2] LI S，SHAO M，FU Y.Cross-view projective dictionary learning for person re-identification[C]//Proceedings of the 24th International Joint Conference on Artificial Intelligence（IJCAI），2015：2156-2161.
[3] LIAO S C，HU Y，ZHU X Y，et al.Person re-identification by local maximal occurrence representation and metric learning[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2015：2197-2206.
[4] SHI Z Y，HOSPEDALES T M，XIANG T.Transferring a semantic representation for person re-identification and search[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2015：4184-4193.
[5] SU C，YANG F，ZHANG S L，et al.Multi-task learning with low rank attribute embedding for person re-identification[C]//IEEE International Conference on Computer Vision（ICCV），2015：3739-3747.
[6] ZHAO D D，WANG H B，YIN H P，et al.Person re-identification by integrating metric learning and support vector machine[J].Signal Processing，2020，166：107-277.
[7] ZHENG W S，GONG S G，XIANG T.Person re-identification by probabilistic relative distance comparison[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2011：649-656.
[8] MA B P，SU Y，JURIE F.Covariance descriptor based on bioinspired features for person re-identification and face verification[J].Image and Vision Computing，2014，32（6/7）：379-390.
[9] FARENZENA M，BAZZANI L，PERINA A，et al.Person re-identification by symmetry-driven accumulation of local features[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），Loris Bazzani，June 3，2014：2360-2367.
[10] MATSUKAWA T，OKABE T，SUZUKI E，et al.Hierarchical Gaussian descriptor for person re-identification[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2016：1363-1372.
[11] GRAY D，TAO H.Viewpoint invariant pedestrian recognition with an ensemble of localized features[C]//European Conference on Computer Vision（ECCV），June 3，2014：262-275.
[12] BOUSMALIS K，TRIGEORGIS G，SILBERMAN N，et al.Domain separation networks[C]//Proceedings of the 29th Conference on Neural Information Processing Systems（NIPS），Barcelona，Spain，Aug 22，2016：343-351.
[13] LIU M Y，BREUEL T，KAUTZ J.Unsupervised image-to-image translation networks[C]//Proceedings of the 31st Conference on Neural Information Processing Systems（NIPS），Long Beach，CA，USA，2017：700-708.
[14] SUN Y F，ZHENG L，DENG W J，et al.Svdnet for pedestrian retrieval[C]//IEEE International Conference on Computer Vision（ICCV），2017：3820-3828.
[15] LI W，LI J Q，ZHU L F.Multiple-shot person re-identification via fair set-collaboration metric learning[J].Neurocomputing，2017，242：15-27.
[16] JIA J R，RUAN Q Q，AN G Y，et al.Multiple metric learning with query adaptive weights and multi-task re-weighting for person re-identification[J].Computer Vision and Image Understanding，2017，160：87-99.
[17] KOSTINGER M，HIRZER M，WOHLHART P，et al.Larger scale metric learning from equivalence constraints[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2012：2288-2295.
[18] LIU X K，WANG H Y，WANG J，et al.Person re-identification by multiple instance metric learning with impostor rejection[J].Pattern Recognition，2017，67：287-298.
[19] SONG K，NIE F P，HAN J W，et al.Parameter free large margin nearest neighbor for distance metric learning[C]//Proceedings of the 31st AAAI Conference on Artificial Intelligence（AAAI），2017：2555-2561.
[20] WANG J，WANG Z，LIANG C，et al.Equidistance constrained metric learning for person re-identification[J].Pattern Recognition，2018，74：38-51.
[21] YANG X，WANG M，TAO D C.Person re-identification with metric learning using privileged information[J].IEEE Transactions on Image Processing，2018，27（2）：791-805.
[22] YI D，LEI Z，LIAO S C，et al.Deep metric learning for person re-identification[C]//Proceedings of 22nd International Conference on Pattern Recognition，2014：34-39.
[23] AN L，CHEN X J，YANG S F，et al.Sparse representation matching for person re-identification[J].Information Sciences，2016，355-356：74-89.
[24] JING X Y，ZHU X K，WU F，et al.Super-resolution person re-identification with semi-coupled low-rank discriminant dictionary learning[J].IEEE Transactions on Image Processing，2017，26（3）：1363-1378.
[25] LI H F，ZHU J T，TAO D P.Asymmetric projection and dictionary learning with listwise and identity consistency constraints for person re-identification[J].IEEE Access，2018，6：37977-37990.
[26] XU W，CHI H Y，ZHOU L，et al.Self-paced least square semi-coupled dictionary learning for person re-identification[C]//IEEE International Conference on Image Processing（ICIP），2017：17-20.
[27] ZHOU Q，ZHENG S B，LING H B，et al.Joint dictionary and metric learning for person re-identification[J].Pattern Recognition，2017，72：196-206.
[28] CHANG Y S，WANG M Y，HE L，et al.Joint deep semantic embedding and metric learning for person re?identification[J].Pattern Recognition Letters，2020，130：1-6.
[29] CHEN S Z，GUO C C，LAI J H.Deep ranking for person re-identification via joint representation learning[J].IEEE Transactions on Image Processing，2016，25（5）：2353-2367.
[30] ZHOU S P，WANG J J，MENG D Y，et al.Deep self-paced learning for person re-identification[J].Pattern Recognition，2018，76：739-751.
[31] NIE F P，HUANG H，CAI X，et al.Efficient and robust feature selection via joint l2，1-norms minimization[C]//Advances in Neural Information Processing Systems，2010：1-9.
[32] DAUBECHIES I，DEFRISE M，DEMOL C.An iterative thresholding algorithm for linear inverse problems with a sparsity constraint[J].Communications on Pure and Applied Mathematics，2004，57（11）：1413-1457.
[33] CAI J F，CANDES E J，SHEN Z W.A singular value thresholding algorithm for matrix completion[J].Siam Journal on Optimization，2008，20（4）：1956-1982.
[34] LEE H，BATTLE A，RAINA R，et al.Efficient sparse coding algorithms[C]//Conference and Workshop on Neural Information Processing Systems（NIPS），2006，19：801-808.
[35] MA L Y，YANG X K，TAO D C.Person re-identification over camera networks using multi-task distance metric learning[J].IEEE Transactions on Image Processing，2014，23（8）：3656-3670.
[36] CHEN D P，YUAN Z J，HUA G，et al.Similarity learning on an explicit polynomial kernel feature map for person re-identification[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2015：1565-1573.
[37] ZHANG Y，LI B H，LU H C，et al.Sample-specific SVM learning for person re-identification[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2016：1278-1287.
[38] TAO D P，GUO Y N，SONG M L，et al.Person re-identification by dual-regularized kiss metric by learning[J].IEEE Transactions on Image Processing，2016，25（6）：2726-2738.
[39] AN L，CHEN X J，YANG S F，et al.Person re-identification by multi-hypergraph fusion[J].IEEE Transactions on Neural Networks and Learning Systems，2017，28（17）：2763-2774.
[40] SUN C，WANG D，LU H C.Person re-identification via distance metric learning with latent variables[J].IEEE Transactions on Image Processing，2017，26（1）：23-34.
[41] CHEN Y C，ZHU X T，ZHENG W S，et al.Person re-identification by camera correlation aware feature augmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2018，40（2）：392-408.
[42] YANG X，WANG M，HONG R，et al.Enhancing person re-identification in a self-trained subspace[J].ACM Transactions on Multimedia Computing，Communications，and Applications，2017，13（3）：1-27.
[43] TAO D P，GUO Y N，YU B S，et al.Deep multi-view feature learning for person re-identification[J].IEEE Transactions on Circuits and Systems for Video Technology，2018，28（10）：2657-2666.
[44] DAI J，ZHANG Y，LU H C，et al.Cross-view semantic projection learning for person re-identification[J].Pattern Recognition，2018，75：63-76.
[45] LEI J J，NIU L J，FU H Z，et al.Person re-identification by semantic region representation and topology constraint[J].IEEE Transactions on Circuits and Systems for Video Technology，2019，29（8）：2453-2466.
[46] HIRZER M，ROTH P M，KOSTINGER M，et al.Relaxed pairwise learned metric for person re-identification[C]//European Conference on Computer Vision（ECCV），2012：780-793.
[47] PAISITKRIANGKRAI S，SHEN C H，HENGEL A V D.Learning to rank in person re-identification with metric ensembles[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2015：1846-1855.
[48] ZHANG L，XIANG T，GONG S G.Learning a discriminative null space for person re-identification[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2016：1239-1248.
[49] CHENG D，GONG Y H，ZHOU S P，et al.Person re-identification by multi-channel parts-based CNN with improved triplet loss function[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2016：1335-1344.
[50] LIN W Y，SHEN Y，YAN J C，et al.Learning correspondence structures for person re-identification[J].IEEE Transactions on Image Processing，2017，26（5）：2438-2453.
[51] LI H F，XU J J，ZHU J T，et al.Top distance regularized projection and dictionary learning for person re-identification[J].Information Sciences，2019，502：472-491.
[52] ZHENG W S，GONG S G，XIANG T.Person re-identification by probabilistic relative distance comparison[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2011：649-656.
[53] XIAO T，LI H S，OUYANG W L，et al.Learning deep feature representations with domain guided dropout for person re-identification[C]//IEEE Conference on Computer Vision and Pattern Recognition（CVPR），2016：1249-1258.
[54] ZHAO C R，WANG X K，WONG W K，et al.Multiple metric learning based on bar-shape descriptor for person re-identification[J].Pattern Recognition，2017，71：218-234.
[55] LI W，LI J Q，ZHU L F.Multiple-shot person re-identification via fair set-collaboration metric learning[J].Neurocomputing，2017，242：15-27.