PLOT: Pseudo-Labeling via Video Object Tracking for Scalable Monocular 3D Object Detection
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909675505582080 |
|---|---|
| author | Lee, Seokyeong Aung, Sithu Choi, Junyong Kim, Seungryong Kim, Ig-Jae Cho, Junghyun |
| author_facet | Lee, Seokyeong Aung, Sithu Choi, Junyong Kim, Seungryong Kim, Ig-Jae Cho, Junghyun |
| contents | Monocular 3D object detection (M3OD) has long faced challenges due to data scarcity caused by high annotation costs and inherent 2D-to-3D ambiguity. Although various weakly supervised methods and pseudo-labeling methods have been proposed to address these issues, they are mostly limited by domain-specific learning or rely solely on shape information from a single observation. In this paper, we propose a novel pseudo-labeling framework that uses only video data and is more robust to occlusion, without requiring a multi-view setup, additional sensors, camera poses, or domain-specific training. Specifically, we explore a technique for aggregating the pseudo-LiDARs of both static and dynamic objects across temporally adjacent frames using object point tracking, enabling 3D attribute extraction in scenarios where 3D data acquisition is infeasible. Extensive experiments demonstrate that our method ensures reliable accuracy and strong scalability, making it a practical and effective solution for M3OD. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_02393 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | PLOT: Pseudo-Labeling via Video Object Tracking for Scalable Monocular 3D Object Detection Lee, Seokyeong Aung, Sithu Choi, Junyong Kim, Seungryong Kim, Ig-Jae Cho, Junghyun Computer Vision and Pattern Recognition Graphics Monocular 3D object detection (M3OD) has long faced challenges due to data scarcity caused by high annotation costs and inherent 2D-to-3D ambiguity. Although various weakly supervised methods and pseudo-labeling methods have been proposed to address these issues, they are mostly limited by domain-specific learning or rely solely on shape information from a single observation. In this paper, we propose a novel pseudo-labeling framework that uses only video data and is more robust to occlusion, without requiring a multi-view setup, additional sensors, camera poses, or domain-specific training. Specifically, we explore a technique for aggregating the pseudo-LiDARs of both static and dynamic objects across temporally adjacent frames using object point tracking, enabling 3D attribute extraction in scenarios where 3D data acquisition is infeasible. Extensive experiments demonstrate that our method ensures reliable accuracy and strong scalability, making it a practical and effective solution for M3OD. |
| title | PLOT: Pseudo-Labeling via Video Object Tracking for Scalable Monocular 3D Object Detection |
| topic | Computer Vision and Pattern Recognition Graphics |
| url | https://arxiv.org/abs/2507.02393 |