3D LiDAR Multi-Object Tracking Using Multi Positive Contrastive Learning and Deep Reinforcement Learning
Due to its precise distance measurement capabilities, 3D LiDAR is a critical sensor in autonomous systems, including autonomous vehicles and self-driving robots. It plays a key role in Multi-Object Tracking (MOT). Current MOT methods typically employ a Tracking-by-Detection(TbD) approach, where obje...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10812747/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832586890237181952 |
|---|---|
| author | Minho Cho Euntai Kim |
| author_facet | Minho Cho Euntai Kim |
| author_sort | Minho Cho |
| collection | DOAJ |
| description | Due to its precise distance measurement capabilities, 3D LiDAR is a critical sensor in autonomous systems, including autonomous vehicles and self-driving robots. It plays a key role in Multi-Object Tracking (MOT). Current MOT methods typically employ a Tracking-by-Detection(TbD) approach, where objects are detected in each frame and matched across frames. However, 3D LiDAR-based tracking faces challenges such as sparsity and occlusion, often leading to ID-switching errors where object identities are incorrectly swapped due to incomplete data. This paper presents a novel 3D LiDAR-based MOT method to address these challenges and enhance tracking accuracy. We propose refining object similarity using contrastive learning, leveraging the distinct shapes of detected objects at varying distances. Additionally, we tackle occlusion issues through reinforcement learning, modeling occlusion dynamics to ensure that re-detected objects retain their original IDs thus improving tracking consistency. Our method is evaluated using the KITTI MOT dataset, demonstrating improved higher-order tracking Accuracy (HOTA) and reduced ID-switching compared to existing 3D LiDAR and camera-LiDAR fusion methods. These findings underscore the effectiveness of our approach across diverse road environments. |
| format | Article |
| id | doaj-art-d275fb847fc741aeaad0c226c5748278 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-d275fb847fc741aeaad0c226c57482782025-01-25T00:00:27ZengIEEEIEEE Access2169-35362025-01-0113124471245710.1109/ACCESS.2024.3521334108127473D LiDAR Multi-Object Tracking Using Multi Positive Contrastive Learning and Deep Reinforcement LearningMinho Cho0https://orcid.org/0000-0001-9026-4744Euntai Kim1https://orcid.org/0000-0002-0975-8390School of Electrical and Electronic Engineering, Yonsei University, Seoul, South KoreaSchool of Electrical and Electronic Engineering, Yonsei University, Seoul, South KoreaDue to its precise distance measurement capabilities, 3D LiDAR is a critical sensor in autonomous systems, including autonomous vehicles and self-driving robots. It plays a key role in Multi-Object Tracking (MOT). Current MOT methods typically employ a Tracking-by-Detection(TbD) approach, where objects are detected in each frame and matched across frames. However, 3D LiDAR-based tracking faces challenges such as sparsity and occlusion, often leading to ID-switching errors where object identities are incorrectly swapped due to incomplete data. This paper presents a novel 3D LiDAR-based MOT method to address these challenges and enhance tracking accuracy. We propose refining object similarity using contrastive learning, leveraging the distinct shapes of detected objects at varying distances. Additionally, we tackle occlusion issues through reinforcement learning, modeling occlusion dynamics to ensure that re-detected objects retain their original IDs thus improving tracking consistency. Our method is evaluated using the KITTI MOT dataset, demonstrating improved higher-order tracking Accuracy (HOTA) and reduced ID-switching compared to existing 3D LiDAR and camera-LiDAR fusion methods. These findings underscore the effectiveness of our approach across diverse road environments.https://ieeexplore.ieee.org/document/10812747/3D LiDARmulti-object trackingcontrastive learningreinforcement learning |
| spellingShingle | Minho Cho Euntai Kim 3D LiDAR Multi-Object Tracking Using Multi Positive Contrastive Learning and Deep Reinforcement Learning IEEE Access 3D LiDAR multi-object tracking contrastive learning reinforcement learning |
| title | 3D LiDAR Multi-Object Tracking Using Multi Positive Contrastive Learning and Deep Reinforcement Learning |
| title_full | 3D LiDAR Multi-Object Tracking Using Multi Positive Contrastive Learning and Deep Reinforcement Learning |
| title_fullStr | 3D LiDAR Multi-Object Tracking Using Multi Positive Contrastive Learning and Deep Reinforcement Learning |
| title_full_unstemmed | 3D LiDAR Multi-Object Tracking Using Multi Positive Contrastive Learning and Deep Reinforcement Learning |
| title_short | 3D LiDAR Multi-Object Tracking Using Multi Positive Contrastive Learning and Deep Reinforcement Learning |
| title_sort | 3d lidar multi object tracking using multi positive contrastive learning and deep reinforcement learning |
| topic | 3D LiDAR multi-object tracking contrastive learning reinforcement learning |
| url | https://ieeexplore.ieee.org/document/10812747/ |
| work_keys_str_mv | AT minhocho 3dlidarmultiobjecttrackingusingmultipositivecontrastivelearninganddeepreinforcementlearning AT euntaikim 3dlidarmultiobjecttrackingusingmultipositivecontrastivelearninganddeepreinforcementlearning |