Research map / 3D vision / Depth and stereo
Bridge-UniPS: Bridging Calibrated Photometric Stereo toward Universal Photometric Stereo
Minzhe Xu, Xiaoyan Liu, YuJie Xing, Qian Chen
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 global confidence feature (ablation) | Bridge-UniPS · DiLiGenT | MAE: 4.7 -> 4.3 | Improved |
| add pixel feature (ablation) | Bridge-UniPS · DiLiGenT | MAE: 4.8 -> 4.3 | Improved |
| add top-K selection of bridge images (ablation) | Bridge-UniPS · DiLiGenT | MAE: 5.6 -> 4.3 | Improved |
| add Dynamic Neural Query module (ablation) | Bridge-UniPS · DiLiGenT | MAE: 6.3 -> 4.3 | Improved |
| add Dynamic Neural Query module (ablation) | Bridge-UniPS · LUCES | MAE: 12.5 -> 7.1 | Improved |
| condition on explicit target-light directions instead of learnable latent tokens (ablation) | Bridge-UniPS · LUCES | MAE: 8.4 -> 7.1 | Improved |
| keep CPS network frozen (ablation) | Bridge-UniPS · DiLiGenT | MAE: 5.4 -> 4.3 | Improved |
| add point-light training data (ablation) | Bridge-UniPS · DiLiGenT | MAE: 4.4 -> 4.3 | No clear effect |
| pair bridge images with consistent target lights instead of shuffled pairing (ablation) | Bridge-UniPS · DiLiGenT | MAE: 44.5 -> 4.3 | Improved |
| use learning-based CPS (Norm-PSN) instead of traditional CPS (Woodham) as downstream (ablation) | Bridge-UniPS · DiLiGenT | MAE: 9.2 -> 4.3 | Improved |
| use Norm-PSN instead of CNN-PS as downstream CPS (ablation) | Bridge-UniPS · DiLiGenT | MAE: 5.4 -> 4.3 | Improved |
| feed bridge images instead of raw inputs into frozen Norm-PSN | Norm-PSN · DiLiGenT | MAE: 6.8 -> 4.3 | Improved |
Closest papers by meaning
- Scalable, Detailed and Mask-Free Universal Photometric Stereo (CVPR 2023 · similarity 0.92)
- Light of Normals: Unified Feature Representation for Universal Photometric Stereo (ICLR 2026 · similarity 0.91)
- Universal Photometric Stereo Network Using Global Lighting Contexts (CVPR 2022 · similarity 0.90)
- Patch-Based Uncalibrated Photometric Stereo Under Natural Illumination (TPAMI 2021 · similarity 0.90)
- Revisiting One-Stage Deep Uncalibrated Photometric Stereo via Fourier Embedding (TPAMI 2025 · similarity 0.89)
- Sparse Views Near Light: A Practical Paradigm for Uncalibrated Point-light Photometric Stereo (CVPR 2024 · similarity 0.87)
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.
