Circle-Based Ratio Loss for Person Reidentification
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 crit...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2020-01-01
|
| Series: | Complexity |
| Online Access: | http://dx.doi.org/10.1155/2020/9860562 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849410469292408832 |
|---|---|
| author | Zhao Yang Jiehao Liu Tie Liu Li Wang Sai Zhao |
| author_facet | Zhao Yang Jiehao Liu Tie Liu Li Wang Sai Zhao |
| author_sort | Zhao Yang |
| collection | DOAJ |
| description | 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. |
| format | Article |
| id | doaj-art-7b8269c4df6f409a9ea3a79b17609586 |
| institution | Kabale University |
| issn | 1076-2787 1099-0526 |
| language | English |
| publishDate | 2020-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Complexity |
| spelling | doaj-art-7b8269c4df6f409a9ea3a79b176095862025-08-20T03:35:06ZengWileyComplexity1076-27871099-05262020-01-01202010.1155/2020/98605629860562Circle-Based Ratio Loss for Person ReidentificationZhao Yang0Jiehao Liu1Tie Liu2Li Wang3Sai Zhao4School of Electronics and Communication Engineering, Guangzhou University, Guangzhou, ChinaSchool of Electronics and Communication Engineering, Guangzhou University, Guangzhou, ChinaSchool of Electronics and Communication Engineering, Guangzhou University, Guangzhou, ChinaSchool of Electronics and Communication Engineering, Guangzhou University, Guangzhou, ChinaSchool of Electronics and Communication Engineering, Guangzhou University, Guangzhou, ChinaPerson 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.http://dx.doi.org/10.1155/2020/9860562 |
| spellingShingle | Zhao Yang Jiehao Liu Tie Liu Li Wang Sai Zhao Circle-Based Ratio Loss for Person Reidentification Complexity |
| title | Circle-Based Ratio Loss for Person Reidentification |
| title_full | Circle-Based Ratio Loss for Person Reidentification |
| title_fullStr | Circle-Based Ratio Loss for Person Reidentification |
| title_full_unstemmed | Circle-Based Ratio Loss for Person Reidentification |
| title_short | Circle-Based Ratio Loss for Person Reidentification |
| title_sort | circle based ratio loss for person reidentification |
| url | http://dx.doi.org/10.1155/2020/9860562 |
| work_keys_str_mv | AT zhaoyang circlebasedratiolossforpersonreidentification AT jiehaoliu circlebasedratiolossforpersonreidentification AT tieliu circlebasedratiolossforpersonreidentification AT liwang circlebasedratiolossforpersonreidentification AT saizhao circlebasedratiolossforpersonreidentification |