EgoFlowNet: Non-Rigid Scene Flow from Point Clouds with Ego-Motion Support
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916310486614016 |
|---|---|
| author | Battrawy, Ramy Schuster, René Stricker, Didier |
| author_facet | Battrawy, Ramy Schuster, René Stricker, Didier |
| contents | Recent weakly-supervised methods for scene flow estimation from LiDAR point clouds are limited to explicit reasoning on object-level. These methods perform multiple iterative optimizations for each rigid object, which makes them vulnerable to clustering robustness. In this paper, we propose our EgoFlowNet - a point-level scene flow estimation network trained in a weakly-supervised manner and without object-based abstraction. Our approach predicts a binary segmentation mask that implicitly drives two parallel branches for ego-motion and scene flow. Unlike previous methods, we provide both branches with all input points and carefully integrate the binary mask into the feature extraction and losses. We also use a shared cost volume with local refinement that is updated at multiple scales without explicit clustering or rigidity assumptions. On realistic KITTI scenes, we show that our EgoFlowNet performs better than state-of-the-art methods in the presence of ground surface points. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_02920 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | EgoFlowNet: Non-Rigid Scene Flow from Point Clouds with Ego-Motion Support Battrawy, Ramy Schuster, René Stricker, Didier Computer Vision and Pattern Recognition Recent weakly-supervised methods for scene flow estimation from LiDAR point clouds are limited to explicit reasoning on object-level. These methods perform multiple iterative optimizations for each rigid object, which makes them vulnerable to clustering robustness. In this paper, we propose our EgoFlowNet - a point-level scene flow estimation network trained in a weakly-supervised manner and without object-based abstraction. Our approach predicts a binary segmentation mask that implicitly drives two parallel branches for ego-motion and scene flow. Unlike previous methods, we provide both branches with all input points and carefully integrate the binary mask into the feature extraction and losses. We also use a shared cost volume with local refinement that is updated at multiple scales without explicit clustering or rigidity assumptions. On realistic KITTI scenes, we show that our EgoFlowNet performs better than state-of-the-art methods in the presence of ground surface points. |
| title | EgoFlowNet: Non-Rigid Scene Flow from Point Clouds with Ego-Motion Support |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2407.02920 |