Lightweight Multimodal Domain Generic Person Reidentification Metric for Person-Following Robots
Recently, person-following robots have been increasingly used in many real-world applications, and they require robust and accurate person identification for tracking. Recent works proposed to use re-identification metrics for identification of the target person; however, these metrics suffer due to...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-01-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/2/813 |
| _version_ | 1797437289797779456 |
|---|---|
| author | Muhammad Adnan Syed Yongsheng Ou Tao Li Guolai Jiang |
| author_facet | Muhammad Adnan Syed Yongsheng Ou Tao Li Guolai Jiang |
| author_sort | Muhammad Adnan Syed |
| collection | DOAJ |
| description | Recently, person-following robots have been increasingly used in many real-world applications, and they require robust and accurate person identification for tracking. Recent works proposed to use re-identification metrics for identification of the target person; however, these metrics suffer due to poor generalization, and due to impostors in nonlinear multi-modal world. This work learns a domain generic person re-identification to resolve real-world challenges and to identify the target person undergoing appearance changes when moving across different indoor and outdoor environments or domains. Our generic metric takes advantage of novel attention mechanism to learn deep cross-representations to address pose, viewpoint, and illumination variations, as well as jointly tackling impostors and style variations the target person randomly undergoes in various indoor and outdoor domains; thus, our generic metric attains higher recognition accuracy of target person identification in complex multi-modal open-set world, and attains 80.73% and 64.44% <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>R</mi><mi>a</mi><mi>n</mi><mi>k</mi></mrow></semantics></math></inline-formula>-1 identification in multi-modal close-set PRID and VIPeR domains, respectively. |
| first_indexed | 2024-03-09T11:16:44Z |
| format | Article |
| id | doaj.art-3f529b6c8bce439c8279e5911f4ff168 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T11:16:44Z |
| publishDate | 2023-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-3f529b6c8bce439c8279e5911f4ff1682023-12-01T00:28:03ZengMDPI AGSensors1424-82202023-01-0123281310.3390/s23020813Lightweight Multimodal Domain Generic Person Reidentification Metric for Person-Following RobotsMuhammad Adnan Syed0Yongsheng Ou1Tao Li2Guolai Jiang3Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, ChinaShenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, ChinaKonka R&D Department, Konka Group Co., Ltd., Shenzhen 518053, ChinaShenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, ChinaRecently, person-following robots have been increasingly used in many real-world applications, and they require robust and accurate person identification for tracking. Recent works proposed to use re-identification metrics for identification of the target person; however, these metrics suffer due to poor generalization, and due to impostors in nonlinear multi-modal world. This work learns a domain generic person re-identification to resolve real-world challenges and to identify the target person undergoing appearance changes when moving across different indoor and outdoor environments or domains. Our generic metric takes advantage of novel attention mechanism to learn deep cross-representations to address pose, viewpoint, and illumination variations, as well as jointly tackling impostors and style variations the target person randomly undergoes in various indoor and outdoor domains; thus, our generic metric attains higher recognition accuracy of target person identification in complex multi-modal open-set world, and attains 80.73% and 64.44% <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>R</mi><mi>a</mi><mi>n</mi><mi>k</mi></mrow></semantics></math></inline-formula>-1 identification in multi-modal close-set PRID and VIPeR domains, respectively.https://www.mdpi.com/1424-8220/23/2/813person re-identificationimpostor resisting metricmulti-modal re-identification metriclightweight domain generic metricpart-wise attention learning |
| spellingShingle | Muhammad Adnan Syed Yongsheng Ou Tao Li Guolai Jiang Lightweight Multimodal Domain Generic Person Reidentification Metric for Person-Following Robots Sensors person re-identification impostor resisting metric multi-modal re-identification metric lightweight domain generic metric part-wise attention learning |
| title | Lightweight Multimodal Domain Generic Person Reidentification Metric for Person-Following Robots |
| title_full | Lightweight Multimodal Domain Generic Person Reidentification Metric for Person-Following Robots |
| title_fullStr | Lightweight Multimodal Domain Generic Person Reidentification Metric for Person-Following Robots |
| title_full_unstemmed | Lightweight Multimodal Domain Generic Person Reidentification Metric for Person-Following Robots |
| title_short | Lightweight Multimodal Domain Generic Person Reidentification Metric for Person-Following Robots |
| title_sort | lightweight multimodal domain generic person reidentification metric for person following robots |
| topic | person re-identification impostor resisting metric multi-modal re-identification metric lightweight domain generic metric part-wise attention learning |
| url | https://www.mdpi.com/1424-8220/23/2/813 |
| work_keys_str_mv | AT muhammadadnansyed lightweightmultimodaldomaingenericpersonreidentificationmetricforpersonfollowingrobots AT yongshengou lightweightmultimodaldomaingenericpersonreidentificationmetricforpersonfollowingrobots AT taoli lightweightmultimodaldomaingenericpersonreidentificationmetricforpersonfollowingrobots AT guolaijiang lightweightmultimodaldomaingenericpersonreidentificationmetricforpersonfollowingrobots |