Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2412.04666 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915711682609152 |
|---|---|
| author | Peng, Kebin Li, Haotang Qi, Zhenyu Chen, Huashan Wang, Zi Zhang, Wei He, Sen Yang, Huanrui Guo, Qing |
| author_facet | Peng, Kebin Li, Haotang Qi, Zhenyu Chen, Huashan Wang, Zi Zhang, Wei He, Sen Yang, Huanrui Guo, Qing |
| contents | State-of-the-art monocular depth estimation (MDE) models often struggle in challenging environments, primarily because they overlook robust physical information. To demonstrate this, we first conduct an empirical study by computing the covariance between a model's prediction error and atmospheric attenuation. We find that the error of existing SOTAs increases with atmospheric attenuation. Based on this finding, we propose PhysDepth, a plug-and-play framework that solves this fragility by infusing physical priors into modern SOTA backbones. PhysDepth incorporates two key components: a Physical Prior Module (PPM) that leverages Rayleigh Scattering theory to extract robust features from the high-SNR red channel, and a physics-derived Red Channel Attenuation Loss (RCA) that enforces model to learn the Beer-Lambert law. Extensive evaluations demonstrate that PhysDepth achieves SOTA accuracy in challenging conditions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_04666 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | PhysDepth: Plug-and-Play Physical Refinement for Monocular Depth Estimation in Challenging Environments Peng, Kebin Li, Haotang Qi, Zhenyu Chen, Huashan Wang, Zi Zhang, Wei He, Sen Yang, Huanrui Guo, Qing Computer Vision and Pattern Recognition State-of-the-art monocular depth estimation (MDE) models often struggle in challenging environments, primarily because they overlook robust physical information. To demonstrate this, we first conduct an empirical study by computing the covariance between a model's prediction error and atmospheric attenuation. We find that the error of existing SOTAs increases with atmospheric attenuation. Based on this finding, we propose PhysDepth, a plug-and-play framework that solves this fragility by infusing physical priors into modern SOTA backbones. PhysDepth incorporates two key components: a Physical Prior Module (PPM) that leverages Rayleigh Scattering theory to extract robust features from the high-SNR red channel, and a physics-derived Red Channel Attenuation Loss (RCA) that enforces model to learn the Beer-Lambert law. Extensive evaluations demonstrate that PhysDepth achieves SOTA accuracy in challenging conditions. |
| title | PhysDepth: Plug-and-Play Physical Refinement for Monocular Depth Estimation in Challenging Environments |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2412.04666 |