{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,2,21]],"date-time":"2025-02-21T02:07:51Z","timestamp":1740103671494,"version":"3.37.3"},"reference-count":48,"publisher":"Wiley","license":[{"start":{"date-parts":[[2020,12,29]],"date-time":"2020-12-29T00:00:00Z","timestamp":1609200000000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["61501177","61902084","YB201712","2018KTSCX174"],"award-info":[{"award-number":["61501177","61902084","YB201712","2018KTSCX174"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100014881","name":"Guangzhou University","doi-asserted-by":"publisher","award":["61501177","61902084","YB201712","2018KTSCX174"],"award-info":[{"award-number":["61501177","61902084","YB201712","2018KTSCX174"]}],"id":[{"id":"10.13039\/501100014881","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100010226","name":"Department of Education of Guangdong Province","doi-asserted-by":"publisher","award":["61501177","61902084","YB201712","2018KTSCX174"],"award-info":[{"award-number":["61501177","61902084","YB201712","2018KTSCX174"]}],"id":[{"id":"10.13039\/501100010226","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Complexity"],"published-print":{"date-parts":[[2020,12,29]]},"abstract":"<jats:p>Person reidentification (re-id) aims to recognize a specific pedestrian from uncrossed surveillance camera views. Most re-id methods perform the retrieval task by comparing the similarity of pedestrian features extracted from deep learning models. Therefore, learning a discriminative feature is critical for person reidentification. Many works supervise the model learning with one or more loss functions to obtain the discriminability of features. Softmax loss is one of the widely used loss functions in re-id. However, traditional softmax loss inherently focuses on the feature separability and fails to consider the compactness of within-class features. To further improve the accuracy of re-id, many efforts are conducted to shrink within-class discrepancy as well as between-class similarity. In this paper, we propose a circle-based ratio loss for person re-identification. Concretely, we normalize the learned features and classification weights to map these vectors in the hypersphere. Then we take the ratio of the maximal intraclass distance and the minimal interclass distance as an objective loss, so the between-class separability and within-class compactness can be optimized simultaneously during the training stage. Finally, with the joint training of an improved softmax loss and the ratio loss, the deep model could mine discriminative pedestrian information and learn robust features for the re-id task. Comprehensive experiments on three re-id benchmark datasets are carried out to illustrate the effectiveness of the proposed method. Specially, 83.12% mAP on Market-1501, 71.66% mAP on DukeMTMC-reID, and 66.26%\/63.24% mAP on CUHK03 labeled\/detected are achieved, respectively.<\/jats:p>","DOI":"10.1155\/2020\/9860562","type":"journal-article","created":{"date-parts":[[2020,12,30]],"date-time":"2020-12-30T02:20:08Z","timestamp":1609294808000},"page":"1-11","source":"Crossref","is-referenced-by-count":1,"title":["Circle-Based Ratio Loss for Person Reidentification"],"prefix":"10.1155","volume":"2020","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-4688-1395","authenticated-orcid":true,"given":"Zhao","family":"Yang","sequence":"first","affiliation":[{"name":"School of Electronics and Communication Engineering, Guangzhou University, Guangzhou, China"},{"name":"Huangpu Research & Graduate School of Guangzhou University, Guangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9835-2797","authenticated-orcid":true,"given":"Jiehao","family":"Liu","sequence":"additional","affiliation":[{"name":"School of Electronics and Communication Engineering, Guangzhou University, Guangzhou, China"},{"name":"Huangpu Research & Graduate School of Guangzhou University, Guangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0968-1645","authenticated-orcid":true,"given":"Tie","family":"Liu","sequence":"additional","affiliation":[{"name":"School of Electronics and Communication Engineering, Guangzhou University, Guangzhou, China"},{"name":"Huangpu Research & Graduate School of Guangzhou University, Guangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9755-5744","authenticated-orcid":true,"given":"Li","family":"Wang","sequence":"additional","affiliation":[{"name":"School of Electronics and Communication Engineering, Guangzhou University, Guangzhou, China"},{"name":"Huangpu Research & Graduate School of Guangzhou University, Guangzhou, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6572-5289","authenticated-orcid":true,"given":"Sai","family":"Zhao","sequence":"additional","affiliation":[{"name":"School of Electronics and Communication Engineering, Guangzhou University, Guangzhou, China"},{"name":"Huangpu Research & Graduate School of Guangzhou University, Guangzhou, China"}]}],"member":"311","reference":[{"key":"1","doi-asserted-by":"publisher","DOI":"10.1109\/tpami.2012.246"},{"first-page":"3594","article-title":"Locally aligned feature transforms across views","author":"W. Li","key":"2"},{"first-page":"2288","article-title":"Large scale metric learning from equivalence constraints","author":"M. K\u00f6stinger","key":"3"},{"first-page":"2197","article-title":"Person re-identification by local maximal occurrence representation and metric learning","author":"S. Liao","key":"4"},{"first-page":"4685","article-title":"ArcFace: additive angular margin loss for deep face recognition","author":"J. Deng","key":"5"},{"first-page":"501","article-title":"Beyond part models: person retrieval with refined part pooling (and a strong convolutional baseline)","author":"Y. Sun","key":"6"},{"article-title":"Person re-identification: past, present, and future","year":"2016","author":"L. Zheng","key":"7"},{"first-page":"1320","article-title":"Beyond triplet loss: a deep quadruplet network for person re-identification","author":"W. Chen","key":"8"},{"first-page":"1735","article-title":"Dimensionality reduction by learning an invariant mapping","author":"R. Hadsell","key":"9"},{"first-page":"791","article-title":"Gated siamese convolutional neural network architecture for human re-identification","author":"R. R. Varior","key":"10"},{"first-page":"22","article-title":"Viewpoint invariant pedestrian recognition with an ensemble of localized features","author":"D. Gray","key":"11"},{"key":"12","doi-asserted-by":"publisher","DOI":"10.1016\/j.jvcir.2019.01.010"},{"article-title":"In defense of the triplet loss for person re-identification","year":"2017","author":"A. Hermans","key":"13"},{"first-page":"6738","article-title":"SphereFace: deep hypersphere embedding for face recognition,","author":"W. Liu","key":"14"},{"first-page":"815","article-title":"FaceNet: a unified embedding for face recognition and clustering","author":"F. Schroff","key":"15"},{"first-page":"770","article-title":"Deep residual learning for image recognition","author":"K. He","key":"16"},{"first-page":"1","article-title":"Going deeper with convolutions","author":"C. Szegedy","key":"17"},{"first-page":"1335","article-title":"Person re-identification by multi-channel parts-based cnn with improved triplet loss function","author":"D. Cheng","key":"18"},{"first-page":"507","article-title":"Large-margin softmax loss for convolutional neural networks","author":"W. Liu","key":"19"},{"first-page":"274","article-title":"Learning discriminative features with multiple granularities for person re-identification","author":"G. Wang","key":"20"},{"first-page":"5265","article-title":"CosFace: large margin cosine loss for deep face recognition","author":"H. Wang","key":"21"},{"key":"22","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2018.03.073"},{"key":"23","doi-asserted-by":"publisher","DOI":"10.1155\/2018\/5940181"},{"first-page":"8514","article-title":"Pyramidal person re-identification via multi-loss dynamic training","author":"F. Zheng","key":"24"},{"key":"25","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2019.01.005"},{"first-page":"542","article-title":"Pose-guided feature alignment for occluded person re-identification","author":"J. Miao","key":"26"},{"key":"27","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2019.12.100"},{"first-page":"6111","article-title":"Self-similarity grouping: a simple unsupervised cross domain adaptation approach for person re-identification,","author":"Y. Fu","key":"28"},{"key":"29","doi-asserted-by":"publisher","DOI":"10.1109\/tmm.2019.2958756"},{"key":"30","doi-asserted-by":"publisher","DOI":"10.1109\/access.2020.3036185"},{"key":"31","article-title":"Pytorch"},{"key":"32","doi-asserted-by":"publisher","DOI":"10.1109\/5.726791"},{"first-page":"1116","article-title":"Scalable person re-identification: a benchmark,","author":"L. Zheng","key":"33"},{"first-page":"3774","article-title":"Unlabeled samples generated by gan improve the person re-identification baseline in vitro","author":"Z. Zheng","key":"34"},{"first-page":"152","article-title":"DeepReID: deep filter pairing neural network for person re-identification","author":"W. Li","key":"35"},{"first-page":"17","article-title":"Performance measures and a data set for multi-target, multi-camera tracking","author":"E. Ristani","key":"36"},{"first-page":"3652","article-title":"Re-ranking person re-identification with k-reciprocal encoding","author":"Z. Zhong","key":"37"},{"article-title":"Random erasing data augmentation","year":"2017","author":"Z. Zhong","key":"38"},{"key":"39","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2019.107036"},{"first-page":"3820","article-title":"SVDNet for pedestrian retrieval","author":"Y. Sun","key":"40"},{"first-page":"2119","article-title":"Attention-aware compositional network for person re-identification","author":"J. Xu","key":"41"},{"first-page":"2590","article-title":"Person re-identification by deep learning multi-scale representations","author":"Y. Chen","key":"42"},{"first-page":"420","article-title":"GLAD: global-local-alignment descriptor for pedestrian retrieval","author":"L. Wei","key":"43"},{"first-page":"2285","article-title":"Harmonious attention network for person re-identification","author":"W. Li","key":"44"},{"first-page":"5363","article-title":"Dual attention marching network for context-aware feature sequence based person re-identification","author":"J. Si","key":"45"},{"first-page":"2265","article-title":"Deep group-shuffling random walk for person re-identification","author":"Y. Shen","key":"46"},{"key":"47","doi-asserted-by":"publisher","DOI":"10.1109\/tcsvt.2018.2873599"},{"first-page":"2109","article-title":"Multi-level factorisation net for person re-identification","author":"X. Chang","key":"48"}],"container-title":["Complexity"],"original-title":[],"language":"en","link":[{"URL":"http:\/\/downloads.hindawi.com\/journals\/complexity\/2020\/9860562.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"http:\/\/downloads.hindawi.com\/journals\/complexity\/2020\/9860562.xml","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"http:\/\/downloads.hindawi.com\/journals\/complexity\/2020\/9860562.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2021,4,23]],"date-time":"2021-04-23T08:01:43Z","timestamp":1619164903000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.hindawi.com\/journals\/complexity\/2020\/9860562\/"}},"subtitle":[],"editor":[{"given":"Atif","family":"Khan","sequence":"additional","affiliation":[]}],"short-title":[],"issued":{"date-parts":[[2020,12,29]]},"references-count":48,"alternative-id":["9860562","9860562"],"URL":"https:\/\/doi.org\/10.1155\/2020\/9860562","relation":{},"ISSN":["1099-0526","1076-2787"],"issn-type":[{"type":"electronic","value":"1099-0526"},{"type":"print","value":"1076-2787"}],"subject":[],"published":{"date-parts":[[2020,12,29]]}}}