A Person Re-Identification Scheme Using Local Multiscale Feature Embedding with Dual Pyramids
In this paper, we propose a new person re-identification scheme that uses dual pyramids to construct and utilize the local multiscale feature embedding that reflects different sizes and shapes of visual feature elements appearing in various areas of a person image. In the dual pyramids, a scale pyra...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-04-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/11/8/3363 |
| _version_ | 1797538389891743744 |
|---|---|
| author | Kwangho Song Yoo-Sung Kim |
| author_facet | Kwangho Song Yoo-Sung Kim |
| author_sort | Kwangho Song |
| collection | DOAJ |
| description | In this paper, we propose a new person re-identification scheme that uses dual pyramids to construct and utilize the local multiscale feature embedding that reflects different sizes and shapes of visual feature elements appearing in various areas of a person image. In the dual pyramids, a scale pyramid reflects the visual feature elements in various sizes and shapes, and a part pyramid selects elements and differently combines them for the feature embedding per each region of the person image. In the experiments, the performance of the cases with and without each pyramid were compared to verify that the proposed scheme has an optimal structure. The state-of-the-art studies known in the field of person re-identification were also compared for accuracy. According to the experimental results, the method proposed in this study showed a maximum of 99.25% Rank-1 accuracy according to the dataset used in the experiments. Based on the same dataset, the accuracy was determined to be about 3.55% higher than the previous studies, which used only person images, and about 1.25% higher than the other studies using additional meta-information besides images of persons. |
| first_indexed | 2024-03-10T12:29:51Z |
| format | Article |
| id | doaj.art-5ae34c08898640bf9593eee8c50fb0b0 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-10T12:29:51Z |
| publishDate | 2021-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-5ae34c08898640bf9593eee8c50fb0b02023-11-21T14:44:02ZengMDPI AGApplied Sciences2076-34172021-04-01118336310.3390/app11083363A Person Re-Identification Scheme Using Local Multiscale Feature Embedding with Dual PyramidsKwangho Song0Yoo-Sung Kim1Department of Information and Communication Engineering, Inha University, Incheon 22212, KoreaDepartment of Information and Communication Engineering, Inha University, Incheon 22212, KoreaIn this paper, we propose a new person re-identification scheme that uses dual pyramids to construct and utilize the local multiscale feature embedding that reflects different sizes and shapes of visual feature elements appearing in various areas of a person image. In the dual pyramids, a scale pyramid reflects the visual feature elements in various sizes and shapes, and a part pyramid selects elements and differently combines them for the feature embedding per each region of the person image. In the experiments, the performance of the cases with and without each pyramid were compared to verify that the proposed scheme has an optimal structure. The state-of-the-art studies known in the field of person re-identification were also compared for accuracy. According to the experimental results, the method proposed in this study showed a maximum of 99.25% Rank-1 accuracy according to the dataset used in the experiments. Based on the same dataset, the accuracy was determined to be about 3.55% higher than the previous studies, which used only person images, and about 1.25% higher than the other studies using additional meta-information besides images of persons.https://www.mdpi.com/2076-3417/11/8/3363person re-identificationdual pyramid structuremultiscale featurepart-wise featurehigh rank-1 accuracy |
| spellingShingle | Kwangho Song Yoo-Sung Kim A Person Re-Identification Scheme Using Local Multiscale Feature Embedding with Dual Pyramids Applied Sciences person re-identification dual pyramid structure multiscale feature part-wise feature high rank-1 accuracy |
| title | A Person Re-Identification Scheme Using Local Multiscale Feature Embedding with Dual Pyramids |
| title_full | A Person Re-Identification Scheme Using Local Multiscale Feature Embedding with Dual Pyramids |
| title_fullStr | A Person Re-Identification Scheme Using Local Multiscale Feature Embedding with Dual Pyramids |
| title_full_unstemmed | A Person Re-Identification Scheme Using Local Multiscale Feature Embedding with Dual Pyramids |
| title_short | A Person Re-Identification Scheme Using Local Multiscale Feature Embedding with Dual Pyramids |
| title_sort | person re identification scheme using local multiscale feature embedding with dual pyramids |
| topic | person re-identification dual pyramid structure multiscale feature part-wise feature high rank-1 accuracy |
| url | https://www.mdpi.com/2076-3417/11/8/3363 |
| work_keys_str_mv | AT kwanghosong apersonreidentificationschemeusinglocalmultiscalefeatureembeddingwithdualpyramids AT yoosungkim apersonreidentificationschemeusinglocalmultiscalefeatureembeddingwithdualpyramids AT kwanghosong personreidentificationschemeusinglocalmultiscalefeatureembeddingwithdualpyramids AT yoosungkim personreidentificationschemeusinglocalmultiscalefeatureembeddingwithdualpyramids |