Research map / Image generation / Fast diffusion sampling
DNF-Intrinsic: Deterministic Noise-Free Diffusion for Indoor Inverse Rendering
Rongjia Zheng, Qing Zhang, Chengjiang Long, Wei-Shi Zheng
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 |
|---|---|---|---|
| use deterministic image-to-intrinsic flow instead of noise-to-intrinsic flow (ablation) | SD3+LoRA · InteriorVerse albedo | PSNR: 18.25 -> 21.95 | Improved |
| use flow matching instead of DDPM v-prediction (ablation) | SD3+Image+LoRA · InteriorVerse albedo | PSNR: 17.62 -> 21.95 | Improved |
| use pre-trained diffusion prior (SD3 DiT) instead of UNet from scratch (ablation) | DNF-Intrinsic · InteriorVerse albedo | PSNR: 17.99 -> 21.95 | Improved |
| use pre-trained diffusion prior instead of DiT from scratch (ablation) | DNF-Intrinsic · InteriorVerse albedo | PSNR: 18.80 -> 21.95 | Improved |
| use SD3 DiT instead of SD2 UNet (ablation) | DNF-Intrinsic · InteriorVerse albedo | PSNR: 19.51 -> 21.95 | Improved |
| fine-tune with LoRA instead of ControlNet (ablation) | DNF-Intrinsic · InteriorVerse albedo | PSNR: 17.13 -> 21.95 | Improved |
| fine-tune with LoRA instead of all parameters (ablation) | DNF-Intrinsic · InteriorVerse albedo | PSNR: 21.30 -> 21.95 | Improved |
| add reconstruction loss (ablation) | DNF-Intrinsic · InteriorVerse albedo | PSNR: 21.03 -> 21.95 | Improved |
| add DNF-Intrinsic vs RGBX on real-world albedo | RGBX · IIW | WHDR: 21.80 -> 20.90 | Improved |
Closest papers by meaning
- Channel-wise Noise Scheduled Diffusion for Inverse Rendering in Indoor Scenes (CVPR 2025 · similarity 0.91)
- Uni-Renderer: Unifying Rendering and Inverse Rendering Via Dual Stream Diffusion (CVPR 2025 · similarity 0.88)
- Diffusion Renderer: Neural Inverse and Forward Rendering with Video Diffusion Models (CVPR 2025 · similarity 0.88)
- Intrinsic Image Diffusion for Indoor Single-view Material Estimation (CVPR 2024 · similarity 0.88)
- Denoising Diffusion via Image-Based Rendering (ICLR 2024 · similarity 0.87)
- A Variational Perspective on Solving Inverse Problems with Diffusion Models (ICLR 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.
