Research map / 3D vision / Depth and stereo
Accurate Depth and Normal Maps From Occlusion-Aware Focal Stack Symmetry
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 |
|---|---|---|---|
| Replace full focal stacks with occlusion-aware partial focal stacks | Focal stack symmetry disparity estimation · Boxes, Cotton, Dino, Sideboard | Pixels deviating more than 0.07 from ground truth (%): 28.12 -> 20.87; 6.83 -> 3.37; 10.01 -> 3.52; 26.80 -> 9.23 | Helped |
| Add confidence-weighted stereo correspondence costs to focal stack symmetry | Proposed disparity estimation · Dots | BadPix(0.07): 37.67 -> 21.61 | Helped |
| Jointly regularize depth and normals instead of using sublabel relaxation alone (ablation) | Proposed depth and normal estimation · Boxes, Cotton, Dino, Sideboard | Mean angular error: 48.55° -> 17.11°; 33.64° -> 9.78°; 43.55° -> 1.68°; 49.59° -> 8.58° | Helped |
| Add normal regularization without joint depth regularization (ablation) | Proposed depth and normal estimation · Boxes, Cotton, Dino, Sideboard | Mean angular error: 36.92° vs 17.11°; 31.54° vs 9.78°; 29.91° vs 1.68°; 65.48° vs 8.58° | Mixed |
| Add depth smoothing only instead of joint depth-normal regularization (ablation) | Proposed depth and normal estimation · Boxes, Cotton, Dino, Sideboard | Mean angular error: 19.67° vs 17.11°; 12.87° vs 9.78°; 2.67° vs 1.68°; 12.49° vs 8.58° | Hurt |
Closest papers by meaning
- RPG360: Robust 360 Depth Estimation with Perspective Foundation Models and Graph Optimization (NeurIPS 2025 · similarity 0.87)
- Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (ICLR 2025 · similarity 0.87)
- DepthMaster: Unified Monocular Depth Estimation for Perspective and Panoramic Images (NeurIPS 2026 · similarity 0.86)
- Zero-shot Depth Completion via Test-time Alignment with Affine-invariant Depth Prior (AAAI 2025 · similarity 0.85)
- DNGaussian: Optimizing Sparse-View 3D Gaussian Radiance Fields with Global-Local Depth Normalization (CVPR 2024 · similarity 0.85)
- A Simple yet Universal Framework for Depth Completion (NeurIPS 2024 · similarity 0.85)
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.