Research map / Image generation / Image restoration
Estimating Reflectance Layer from a Single Image: Integrating Reflectance Guidance and Shadow/Specular Aware Learning
Yeying Jin, Ruoteng Li, Wenhan Yang, Robby T. Tan
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 |
|---|---|---|---|
| Use the two-stage reflectance guidance and S-Aware method | Ours · IIW | WHDR%: 14.45 -> 17.97 | Hurt |
| Use the two-stage reflectance guidance and S-Aware method | Ours · ShapeNet intrinsic | si-MSE R: 1.85 -> 0.63 | Improved |
| Use the two-stage reflectance guidance and S-Aware method | Ours · ShapeNet intrinsic | si-MSE S: 1.08 -> 2.00 | Hurt |
| Use the two-stage reflectance guidance and S-Aware method | Ours · ShapeNet intrinsic | si-MSE Avg.: 1.47 -> 1.31 | Improved |
| Use the two-stage reflectance guidance and S-Aware method | Ours · ShapeNet intrinsic | si-LMSE Total: 4.65 -> 4.14 | Improved |
| Add stage1 reflectance guidance instead of bypassing it (ablation) | Ours · ShapeNet intrinsic | si-MSE R: 0.85 -> 0.63 | Improved |
| Add stage1 reflectance guidance instead of bypassing it (ablation) | Ours · ShapeNet intrinsic | si-MSE Avg.: 1.51 -> 1.31 | Improved |
| Add stage2 refinement instead of bypassing it (ablation) | Ours · ShapeNet intrinsic | si-MSE Avg.: 1.57 -> 1.31 | Improved |
| Add the S-Classifier and classification loss (ablation) | S-Aware network · ShapeNet intrinsic | si-MSE Avg.: 1.42 -> 1.31 | Improved |
| Add the shadow-free loss (ablation) | first-stage reflectance guidance · ShapeNet intrinsic | si-MSE Avg.: 1.40 -> 1.31 | Improved |
| Add the specular-free loss (ablation) | first-stage reflectance guidance · ShapeNet intrinsic | si-MSE Avg.: 1.35 -> 1.31 | Improved |
Closest papers by meaning
- Reflectance and Natural Illumination from Single-Material Specular Objects Using Deep Learning (TPAMI 2017 · similarity 0.89)
- Reflectance Adaptive Filtering Improves Intrinsic Image Estimation (CVPR 2017 · similarity 0.88)
- Diffusion Reflectance Map: Single-Image Stochastic Inverse Rendering of Illumination and Reflectance (CVPR 2024 · similarity 0.85)
- S^3-Face: SSS-Compliant Facial Reflectance Estimation via Diffusion Priors (CVPR 2025 · similarity 0.84)
- SpecNeRF: Gaussian Directional Encoding for Specular Reflections (CVPR 2024 · similarity 0.84)
- When Shadow Removal Meets Intrinsic Image Decomposition: A Joint Learning Framework Using Unpaired Data (AAAI 2025 · similarity 0.84)
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.
