Research map / Image generation / Image restoration
MatE: Material Extraction from Single-Image via Geometric Prior
Zeyu Zhang, Wei Zhai, Jian Yang, Yang Cao
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 rotation alignment and perspective rectification (ablation) | MatE · Real-world material extraction (attribute-averaged) | LPIPS: 0.464 -> 0.418 | Improved |
| add rotation alignment (with rectification) (ablation) | MatE · Real-world material extraction (attribute-averaged) | LPIPS: 0.428 -> 0.418 | Improved |
| add perspective rectification (with rotation alignment) (ablation) | MatE · Real-world material extraction (attribute-averaged) | LPIPS: 0.435 -> 0.418 | Improved |
| add rotation alignment (with rectification) (ablation) | MatE · Real-world material extraction (attribute-averaged) | SSIM: 0.396 -> 0.377 | Hurt |
| add perspective rectification (with rotation alignment) (ablation) | MatE · Real-world material extraction (attribute-averaged) | SSIM: 0.391 -> 0.377 | Hurt |
| add rotation alignment and perspective rectification with the new dataset (ablation) | MaterialPicker* · Real-world material extraction (attribute-averaged) | LPIPS: 0.526 -> 0.425 | Improved |
| use proposed conditioning instead of CLIP feature injection (ablation) | MatE · Real-world material extraction (attribute-averaged) | LPIPS: 0.674 -> 0.418 | Improved |
| use proposed conditioning instead of latent concatenation (ablation) | MatE · Real-world material extraction (attribute-averaged) | LPIPS: 0.529 -> 0.418 | Improved |
| use proposed conditioning instead of ControlNet-style addition (ablation) | MatE · Real-world material extraction (attribute-averaged) | LPIPS: 0.447 -> 0.418 | Improved |
| use proposed conditioning instead of ControlNet-style addition (ablation) | MatE · Real-world material extraction (attribute-averaged) | CLIP score: 0.723 -> 0.723 | No clear effect |
| use MatE instead of MaterialPicker* | MatE · Real-world Albedo | LPIPS: 0.554 -> 0.445 | Improved |
| use MatE instead of MaterialPicker* | MatE · Real-world Normal | SSIM: 0.317 -> 0.248 | Hurt |
Closest papers by meaning
- Material Palette: Extraction of Materials from a Single Image (CVPR 2024 · similarity 0.93)
- TexGaussian: Generating High-quality PBR Material via Octree-based 3D Gaussian Splatting (CVPR 2025 · similarity 0.89)
- MAGE : Single Image to Material-Aware 3D via the Multi-View G-Buffer Estimation Model (CVPR 2025 · similarity 0.88)
- IDArb: Intrinsic Decomposition for Arbitrary Number of Input Views and Illuminations (ICLR 2025 · similarity 0.88)
- MaterialRefGS: Reflective Gaussian Splatting with Multi-view Consistent Material Inference (NeurIPS 2025 · similarity 0.88)
- Material Anything: Generating Materials for Any 3D Object via Diffusion (CVPR 2025 · similarity 0.88)
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.
