Towards Sharper Object Boundaries in Self-Supervised Depth Estimation
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_ | 1866915623556087808 |
|---|---|
| author | Cecille, Aurélien Duffner, Stefan Davoine, Franck Agier, Rémi Neveu, Thibault |
| author_facet | Cecille, Aurélien Duffner, Stefan Davoine, Franck Agier, Rémi Neveu, Thibault |
| contents | Accurate monocular depth estimation is crucial for 3D scene understanding, but existing methods often blur depth at object boundaries, introducing spurious intermediate 3D points. While achieving sharp edges usually requires very fine-grained supervision, our method produces crisp depth discontinuities using only self-supervision. Specifically, we model per-pixel depth as a mixture distribution, capturing multiple plausible depths and shifting uncertainty from direct regression to the mixture weights. This formulation integrates seamlessly into existing pipelines via variance-aware loss functions and uncertainty propagation. Extensive evaluations on KITTI and VKITTIv2 show that our method achieves up to 35% higher boundary sharpness and improves point cloud quality compared to state-of-the-art baselines. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_15987 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Towards Sharper Object Boundaries in Self-Supervised Depth Estimation Cecille, Aurélien Duffner, Stefan Davoine, Franck Agier, Rémi Neveu, Thibault Computer Vision and Pattern Recognition Artificial Intelligence Robotics Accurate monocular depth estimation is crucial for 3D scene understanding, but existing methods often blur depth at object boundaries, introducing spurious intermediate 3D points. While achieving sharp edges usually requires very fine-grained supervision, our method produces crisp depth discontinuities using only self-supervision. Specifically, we model per-pixel depth as a mixture distribution, capturing multiple plausible depths and shifting uncertainty from direct regression to the mixture weights. This formulation integrates seamlessly into existing pipelines via variance-aware loss functions and uncertainty propagation. Extensive evaluations on KITTI and VKITTIv2 show that our method achieves up to 35% higher boundary sharpness and improves point cloud quality compared to state-of-the-art baselines. |
| title | Towards Sharper Object Boundaries in Self-Supervised Depth Estimation |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Robotics |
| url | https://arxiv.org/abs/2509.15987 |