Estimating 2D Camera Motion with Hybrid Motion Basis
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913966020624384 |
|---|---|
| author | Li, Haipeng Zhou, Tianhao Yang, Zhanglei Wu, Yi Chen, Yan Mao, Zijing Cheng, Shen Zeng, Bing Liu, Shuaicheng |
| author_facet | Li, Haipeng Zhou, Tianhao Yang, Zhanglei Wu, Yi Chen, Yan Mao, Zijing Cheng, Shen Zeng, Bing Liu, Shuaicheng |
| contents | Estimating 2D camera motion is a fundamental computer vision task that models the projection of 3D camera movements onto the 2D image plane. Current methods rely on either homography-based approaches, limited to planar scenes, or meshflow techniques that use grid-based local homographies but struggle with complex non-linear transformations. A key insight of our work is that combining flow fields from different homographies creates motion patterns that cannot be represented by any single homography. We introduce CamFlow, a novel framework that represents camera motion using hybrid motion bases: physical bases derived from camera geometry and stochastic bases for complex scenarios. Our approach includes a hybrid probabilistic loss function based on the Laplace distribution that enhances training robustness. For evaluation, we create a new benchmark by masking dynamic objects in existing optical flow datasets to isolate pure camera motion. Experiments show CamFlow outperforms state-of-the-art methods across diverse scenarios, demonstrating superior robustness and generalization in zero-shot settings. Code and datasets are available at our project page: https://lhaippp.github.io/CamFlow/. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_22480 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Estimating 2D Camera Motion with Hybrid Motion Basis Li, Haipeng Zhou, Tianhao Yang, Zhanglei Wu, Yi Chen, Yan Mao, Zijing Cheng, Shen Zeng, Bing Liu, Shuaicheng Computer Vision and Pattern Recognition Estimating 2D camera motion is a fundamental computer vision task that models the projection of 3D camera movements onto the 2D image plane. Current methods rely on either homography-based approaches, limited to planar scenes, or meshflow techniques that use grid-based local homographies but struggle with complex non-linear transformations. A key insight of our work is that combining flow fields from different homographies creates motion patterns that cannot be represented by any single homography. We introduce CamFlow, a novel framework that represents camera motion using hybrid motion bases: physical bases derived from camera geometry and stochastic bases for complex scenarios. Our approach includes a hybrid probabilistic loss function based on the Laplace distribution that enhances training robustness. For evaluation, we create a new benchmark by masking dynamic objects in existing optical flow datasets to isolate pure camera motion. Experiments show CamFlow outperforms state-of-the-art methods across diverse scenarios, demonstrating superior robustness and generalization in zero-shot settings. Code and datasets are available at our project page: https://lhaippp.github.io/CamFlow/. |
| title | Estimating 2D Camera Motion with Hybrid Motion Basis |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2507.22480 |