Research map / 3D vision / Depth and stereo
Zero-shot Depth from Defocus
Yiming Zuo, Hongyu Wen, Venkat Subramanian, Patrick Chen, Karhan Kayan, Mario Bijelic, Felix Heide, Jia Deng
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add stack attention layer (ablation) | FOSSA ViT-S · ZEDD val | delta_1.05: 0.092 -> 0.445 | Improved |
| add focus distance embedding (ablation) | FOSSA ViT-S · ZEDD val | delta_1.05: 0.090 -> 0.445 | Improved |
| collapse at L1=4 instead of L1=2 (ablation) | FOSSA ViT-S · ZEDD val | delta_1.05: 0.372 -> 0.445 | Improved |
| collapse at L1=6 instead of L1=4 (ablation) | FOSSA ViT-S · ZEDD val | delta_1.05: 0.445 -> 0.524 | Improved |
| collapse in feature extractor instead of fusing in DPT head (ablation) | FOSSA ViT-S · ZEDD val | delta_1.05: 0.061 -> 0.445 | Improved |
| add random blur kernel (PSF shape and f-number) during training (ablation) | FOSSA ViT-S · ZEDD val | delta_1.25: drops of 21.12% | Improved |
| add random focus distance distribution during training (ablation) | FOSSA ViT-S · ZEDD val | delta_1.25: drops of 39.78% | Improved |
| retrain DFF-DFV with FOSSA training data (ablation) | DFF-DFV · DIODE | delta_1.25: 11.92% improvement | Improved |
| use FOSSA ViT-B over DFF-DFV retrained on same data (ablation) | FOSSA · DIODE | AbsRel: 1.331 -> 0.160 | Improved |
| use FOSSA instead of DepthPro (ablation) | FOSSA ViT-B · ZEDD test | AbsRel: 0.201 -> 0.089 | Improved |
| use FOSSA instead of DepthPro (ablation) | FOSSA ViT-S · Infinigen Defocus | AbsRel: 0.176 -> 0.085 | Improved |
| finetune on DDFF instead of zero-shot (ablation) | FOSSA ViT-B · DDFF | MSE (x1e-4): 12.9 -> 2.8 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- MoGe-2 — MoGe-2: Accurate Monocular Geometry with Metric Scale and Sharp Details (NeurIPS 2025) · better on 4 of 4 shared results
- DepthPro — Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (ICLR 2025) · better on 4 of 4 shared results
- DualFocus — DualFocus: Depth from Focus with Spatio-Focal Dual Variational Constraints (NeurIPS 2025) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Repurposing Marigold for Zero-Shot Metric Depth Estimation via Defocus Blur Cues (NeurIPS 2025 · similarity 0.89)
- Zero-shot Depth Completion via Test-time Alignment with Affine-invariant Depth Prior (AAAI 2025 · similarity 0.88)
- Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (ICLR 2025 · similarity 0.88)
- DepthMaster: Unified Monocular Depth Estimation for Perspective and Panoramic Images (NeurIPS 2026 · similarity 0.86)
- Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data (CVPR 2024 · similarity 0.86)
- RPG360: Robust 360 Depth Estimation with Perspective Foundation Models and Graph Optimization (NeurIPS 2025 · similarity 0.86)
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.
