基于卷积神经网络和Transformer的视频行人再识别
TP391; 为了解决视频行人再识别领域仅使用卷积神经网络进行行人特征提取效果不佳的问题,提出一种基于卷积神经网络和Transformer的ResTNet(ResNet and Transformer network)网络模型.ResTNet利用ResNet50网络得到局部特征,令中间层输出作为Transformer的先验知识输入.在Transformer分支中不断缩小特征图尺寸,扩大感受野,充分挖掘局部特征之间的关系,生成行人的全局特征,同时利用移位窗口方法减少模型计算量.在大规模MARS数据集上,Rank-1和mAP分别达到 86.8％和 80.3％,比基准分别增加了 3.8％和 3.3％...
Saved in:
| Published in | 河南理工大学学报（自然科学版） Vol. 42; no. 6; pp. 149 - 156 |
|---|---|
| Main Authors | , , , |
| Format | Journal Article |
| Language | Chinese |
| Published |
河南理工大学 机械与动力工程学院,河南 焦作 454000
01.11.2023
|
| Subjects | |
| Online Access | Get full text |
Cover
Loading…
| Abstract | TP391; 为了解决视频行人再识别领域仅使用卷积神经网络进行行人特征提取效果不佳的问题,提出一种基于卷积神经网络和Transformer的ResTNet(ResNet and Transformer network)网络模型.ResTNet利用ResNet50网络得到局部特征,令中间层输出作为Transformer的先验知识输入.在Transformer分支中不断缩小特征图尺寸,扩大感受野,充分挖掘局部特征之间的关系,生成行人的全局特征,同时利用移位窗口方法减少模型计算量.在大规模MARS数据集上,Rank-1和mAP分别达到 86.8％和 80.3％,比基准分别增加了 3.8％和 3.3％,在 2个小规模数据集上也取得了良好效果.在几大数据集上的大量实验表明,本文方法能增强行人识别的鲁棒性,有效提高行人再识别的准确率. |
|---|---|
| AbstractList | TP391; 为了解决视频行人再识别领域仅使用卷积神经网络进行行人特征提取效果不佳的问题,提出一种基于卷积神经网络和Transformer的ResTNet(ResNet and Transformer network)网络模型.ResTNet利用ResNet50网络得到局部特征,令中间层输出作为Transformer的先验知识输入.在Transformer分支中不断缩小特征图尺寸,扩大感受野,充分挖掘局部特征之间的关系,生成行人的全局特征,同时利用移位窗口方法减少模型计算量.在大规模MARS数据集上,Rank-1和mAP分别达到 86.8％和 80.3％,比基准分别增加了 3.8％和 3.3％,在 2个小规模数据集上也取得了良好效果.在几大数据集上的大量实验表明,本文方法能增强行人识别的鲁棒性,有效提高行人再识别的准确率. |
| Abstract_FL | To solve the problem of poor effect of person feature extraction using only convolutional neural network in the field of video person re-identification,a network model ResTNet(ResNet and Transformer Network)based on convolutional neural network and Transformer was proposed.ResNet50 network was used to obtain local features and the output of its middle layer was input to Transformer as prior knowledge in ResTNet.In the Transformer branch,the size of the feature map was continuously reduced,the field of per-ception was expanded,and the relationship among local features was fully explored to generate the global features of pedestrians,while the model computation was decreased with the shift window method.The Rank-1 and mAP on the large-scale MARS dataset reached 86.8％and 80.3％,respectively,which were 3.8％and 3.3％higher than the benchmark.Meanwhile,excellent performance was also achieved on the two small-scale datasets.In this paper,not only the Transformer model was successfully applied to the field of video person re-identification,but also extensive experiments on several large datasets showed that the proposed ResTNet network could enhance the robustness of the recognition and improve the accuracy of person re-identification effectively. |
| Author | 牛东杰 赵彦如 孙东红 杨蕙萌 |
| AuthorAffiliation | 河南理工大学 机械与动力工程学院,河南 焦作 454000 |
| AuthorAffiliation_xml | – name: 河南理工大学 机械与动力工程学院,河南 焦作 454000 |
| Author_FL | YANG Huimeng NIU Dongjie SUN Donghong ZHAO Yanru |
| Author_FL_xml | – sequence: 1 fullname: ZHAO Yanru – sequence: 2 fullname: NIU Dongjie – sequence: 3 fullname: SUN Donghong – sequence: 4 fullname: YANG Huimeng |
| Author_xml | – sequence: 1 fullname: 赵彦如 – sequence: 2 fullname: 牛东杰 – sequence: 3 fullname: 孙东红 – sequence: 4 fullname: 杨蕙萌 |
| BookMark | eNo9j71Lw0AAxW-oYK39LxxcEu8juY9FkOIXFFzqXO6SS2nUCySI1VmCQ1upo4iiIB0cOgop_jcNF_8LA4q84fF-w3u8DdAwidEAbCHoIoo43YndwJwN68CIIxhnLoYYIQwhIg3Q_MfroJ1lQwVRLUIgboLd8qVYFdNy8mnnC_v-bJf39mtml0_lw7iXSpNFSXqhU_t4W83z77dZ9TpeFUWZT6pFXt59bIK1SJ5nuv3nLXB6sN_rHDndk8Pjzl7XyRD0haP90NOUKBQETEGlMOdMhFxTj0gOhR-EnHmRj5WgTEAUcqIw9jzlh5IyrgRpge3f3itpImkG_Ti5TE292I9vBqPrkar_EkghEuQHUe5h3w |
| ClassificationCodes | TP391 |
| ContentType | Journal Article |
| Copyright | Copyright © Wanfang Data Co. Ltd. All Rights Reserved. |
| Copyright_xml | – notice: Copyright © Wanfang Data Co. Ltd. All Rights Reserved. |
| DBID | 2B. 4A8 92I 93N PSX TCJ |
| DOI | 10.16186/j.cnki.1673-9787.2021120013 |
| DatabaseName | Wanfang Data Journals - Hong Kong WANFANG Data Centre Wanfang Data Journals 万方数据期刊 - 香港版 China Online Journals (COJ) China Online Journals (COJ) |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| DocumentTitle_FL | Video person re-identification based on convolutional neural network and Transformer |
| EndPage | 156 |
| ExternalDocumentID | jzgxyxb202306019 |
| GrantInformation_xml | – fundername: (国家自然科学基金); (河南省科技攻关计划项目); (河南理工大学光电传感与智能测控河南省工程实验室开放课题) funderid: (国家自然科学基金); (河南省科技攻关计划项目); (河南理工大学光电传感与智能测控河南省工程实验室开放课题) |
| GroupedDBID | 2B. 4A8 92I 93N ALMA_UNASSIGNED_HOLDINGS GROUPED_DOAJ PSX TCJ |
| ID | FETCH-LOGICAL-s1059-e5d4e63b1cc7b0bb28879d8e643a8095cd874f52b967901d83b2244b5da678b93 |
| ISSN | 1673-9787 |
| IngestDate | Thu May 29 04:07:54 EDT 2025 |
| IsPeerReviewed | false |
| IsScholarly | true |
| Issue | 6 |
| Keywords | 卷积神经网络 Transformer global fea-ture 视频行人再识别 video person re-identification convolutional neural network local feature 全局特征 局部特征 |
| Language | Chinese |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-s1059-e5d4e63b1cc7b0bb28879d8e643a8095cd874f52b967901d83b2244b5da678b93 |
| PageCount | 8 |
| ParticipantIDs | wanfang_journals_jzgxyxb202306019 |
| PublicationCentury | 2000 |
| PublicationDate | 2023-11-01 |
| PublicationDateYYYYMMDD | 2023-11-01 |
| PublicationDate_xml | – month: 11 year: 2023 text: 2023-11-01 day: 01 |
| PublicationDecade | 2020 |
| PublicationTitle | 河南理工大学学报（自然科学版） |
| PublicationTitle_FL | Journal of Henan Polytechnic University(Natural Science) |
| PublicationYear | 2023 |
| Publisher | 河南理工大学 机械与动力工程学院,河南 焦作 454000 |
| Publisher_xml | – name: 河南理工大学 机械与动力工程学院,河南 焦作 454000 |
| SSID | ssib010103302 ssj0003314027 ssib006704847 ssib051373601 ssib036434603 ssib031741050 ssib005319289 ssib002423915 ssib011070700 ssib008679455 ssib006595874 |
| Score | 2.3417714 |
| Snippet | TP391; 为了解决视频行人再识别领域仅使用卷积神经网络进行行人特征提取效果不佳的问题,提出一种基于卷积神经网络和Transformer的ResTNet(ResNet and Transformer... |
| SourceID | wanfang |
| SourceType | Aggregation Database |
| StartPage | 149 |
| Title | 基于卷积神经网络和Transformer的视频行人再识别 |
| URI | https://d.wanfangdata.com.cn/periodical/jzgxyxb202306019 |
| Volume | 42 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwnR1NaxQxdCgV1IsoKn5ToTmVqTOZJJNchMl0liLoqYXeynxt_YAV7BZqz1I8tEo9iigK0oOHHoUt_psuW_-F72XTmWhFamEZsm9e3ufM5L2QvHjeNA8LGtNa-KXoch9rSPl5HtY-y7ulrHhZ0Qj3Oz98JOYX2YMlvjQxec5ZtbTWL2bLjb_uKzmNVwEGfsVdsv_h2YYoAKAN_oUreBiuJ_IxyThRHaITkjG8ygwhco7omGQxSeDXMQ1Ay7ChNZEGoueICi1EpYYOJTJdOApjwZFwUwFJRjKJlKQgmSIJNf0AEgK6ZYv8OSIAZ7zVMcgAgbZ2o1-SCaIp0ZEVUxkxVWDxtZEUGglDjtiYI4lwGsAiMTigdIr0gZ2EXgnSAVG1sHqjkE0vuKUM8rhXM_WBvTXHHzJvWAET2qKYzkobVaWxlSAKLBy0KIaRUg5KjDZJHCqmT2LkVdzgSqN36s660MhuPzx6T05nrxnDLjVuEegxHVjR7PORGElMAxVrCMUIl9qxnEJRgQJNj4syY22OeMw8UOkMFloMAmegE3Hkq9gGO3YkZNR5491hLRyXlbURUjguBX9s8MWjF8zoW_aePZltOMyC_SCqx2SjDTqapaBPN1bWX64X1OTCARbwPUMh46PO7IiNJPEog9-GDuoclIBlMaWTyYgYhiYnU8a6kqzdgh3i8SZO5TucBQmcSpUQVeP65_Y_BO5MtAsWeBjFkbArNjDoi6KQBeaA6Ebvs970kVXu_cMmZrdgr5v3VpzAduGid8FmpFPJ-PNyyZvYeHzZuz_8NDgYvBlufx_t7o2-fhztvx392Bntfxi-23I-EqP3rw53N39-2Tn8vHUwGAw3tw_3Noevv13xFjvZQjrv26NW_FVMsPyaV6wWURGWZVwERUEh9lCVrEHtXEIWVlZg3C6nhQI7BmElowJif1bwKodot1DRVW-y97xXX_OmiorHNauiuMsoi0qZq1h1OadBqVQtWXXdu2v1Xbaf0tXlPx-CGyfAuemdb1_MW95k_8VafRsShH5xxzw6vwBwZcJK |
| linkProvider | Directory of Open Access Journals |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=%E5%9F%BA%E4%BA%8E%E5%8D%B7%E7%A7%AF%E7%A5%9E%E7%BB%8F%E7%BD%91%E7%BB%9C%E5%92%8CTransformer%E7%9A%84%E8%A7%86%E9%A2%91%E8%A1%8C%E4%BA%BA%E5%86%8D%E8%AF%86%E5%88%AB&rft.jtitle=%E6%B2%B3%E5%8D%97%E7%90%86%E5%B7%A5%E5%A4%A7%E5%AD%A6%E5%AD%A6%E6%8A%A5%EF%BC%88%E8%87%AA%E7%84%B6%E7%A7%91%E5%AD%A6%E7%89%88%EF%BC%89&rft.au=%E8%B5%B5%E5%BD%A6%E5%A6%82&rft.au=%E7%89%9B%E4%B8%9C%E6%9D%B0&rft.au=%E5%AD%99%E4%B8%9C%E7%BA%A2&rft.au=%E6%9D%A8%E8%95%99%E8%90%8C&rft.date=2023-11-01&rft.pub=%E6%B2%B3%E5%8D%97%E7%90%86%E5%B7%A5%E5%A4%A7%E5%AD%A6+%E6%9C%BA%E6%A2%B0%E4%B8%8E%E5%8A%A8%E5%8A%9B%E5%B7%A5%E7%A8%8B%E5%AD%A6%E9%99%A2%2C%E6%B2%B3%E5%8D%97+%E7%84%A6%E4%BD%9C+454000&rft.issn=1673-9787&rft.volume=42&rft.issue=6&rft.spage=149&rft.epage=156&rft_id=info:doi/10.16186%2Fj.cnki.1673-9787.2021120013&rft.externalDocID=jzgxyxb202306019 |
| thumbnail_s | http://utb.summon.serialssolutions.com/2.0.0/image/custom?url=http%3A%2F%2Fwww.wanfangdata.com.cn%2Fimages%2FPeriodicalImages%2Fjzgxyxb%2Fjzgxyxb.jpg |