Research map / 3D vision / Depth and stereo
P3Depth: Monocular Depth Estimation With a Piecewise Planarity Prior
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use plane coefficients instead of direct depth prediction, without offsets (ablation) | P3Depth · NYU Depth-v2 | A.Rel: 0.142 -> 0.144 | Hurt |
| Add offset vectors to plane-coefficient prediction (ablation) | P3Depth · NYU Depth-v2 | A.Rel: 0.144 -> 0.116 | Helped |
| Add offset vectors to direct depth prediction (ablation) | P3Depth · NYU Depth-v2 | A.Rel: 0.142 -> 0.140 | Helped |
| Add the guidance module to plane coefficients with offsets (ablation) | P3Depth · NYU Depth-v2 | A.Rel: 0.116 -> 0.115 | Helped |
| Increase maximum offset length from τ=0.1 to τ=0.3 (ablation) | P3Depth · NYU Depth-v2 | A.Rel: 0.115 -> 0.116 | Hurt |
| Add cascaded offset refinement to plane coefficients with offsets (ablation) | P3Depth · NYU Depth-v2 | A.Rel: 0.116 -> 0.113 | Helped |
| Add cascaded offset refinement to direct depth with offsets (ablation) | P3Depth · NYU Depth-v2 | A.Rel: 0.140 -> 0.134 | Helped |
| Add the mean plane loss to plane coefficients, offsets, and refinement (ablation) | P3Depth · NYU Depth-v2 | A.Rel: 0.113 -> 0.109 | Helped |
| Add the mean plane loss to the full guidance-based model (ablation) | P3Depth · NYU Depth-v2 | A.Rel: 0.109 -> 0.104 | Helped |
| Use P3Depth instead of Long et al. | NYU Depth-v2 Eigen test · NYU Depth-v2 | Log10: 0.044 -> 0.043 | Helped |
| Use P3Depth instead of Ranftl et al. | NYU Depth-v2 Eigen test · NYU Depth-v2 | RMSE: 0.357 -> 0.356 | Helped |
| Use P3Depth instead of the prior best method | KITTI Garg split · KITTI Garg split | RMSE: 1.925 -> 1.651 | Helped |
Closest papers by meaning
- GeoDepth: From Point-to-Depth to Plane-to-Depth Modeling for Self-Supervised Monocular Depth Estimation (CVPR 2025 · similarity 0.91)
- A Simple yet Universal Framework for Depth Completion (NeurIPS 2024 · similarity 0.90)
- Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (ICLR 2025 · similarity 0.90)
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.89)
- MonoInstance: Enhancing Monocular Priors via Multi-view Instance Alignment for Neural Rendering and Reconstruction (CVPR 2025 · similarity 0.89)
- UniDepth: Universal Monocular Metric Depth Estimation (CVPR 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.