Research map / 3D vision / NeRF and radiance fields
DiffRF: Rendering-Guided 3D Radiance Field Diffusion
Norman Müller, Yawar Siddiqui, Lorenzo Porzi, Samuel Rota Bulò, Peter Kontschieder, Matthias Nießner
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 2D rendering supervision (ablation) | DiffRF · PhotoShape Chairs | FID: 18.27 -> 15.95 | Improved |
| Add 2D rendering supervision (ablation) | DiffRF · PhotoShape Chairs | KID: 9.263 -> 7.935 | Improved |
| Add 2D rendering supervision (ablation) | DiffRF · PhotoShape Chairs | COV: 59.20 -> 58.93 | Hurt |
| Add 2D rendering supervision (ablation) | DiffRF · PhotoShape Chairs | MMD: 4.543 -> 4.416 | Improved |
| Add 2D rendering supervision (ablation) | DiffRF · ABO Tables | FID: 35.89 -> 27.06 | Improved |
| Add 2D rendering supervision (ablation) | DiffRF · ABO Tables | KID: 13.94 -> 10.03 | Improved |
| Add 2D rendering supervision (ablation) | DiffRF · ABO Tables | COV: 63.46 -> 61.54 | Hurt |
| Add 2D rendering supervision (ablation) | DiffRF · ABO Tables | MMD: 8.013 -> 7.610 | Improved |
| Use DiffRF rather than EG3D for unconditional synthesis | PhotoShape Chairs · PhotoShape Chairs | FID: 16.54 -> 15.95 | Improved |
| Use DiffRF rather than EG3D for unconditional synthesis | ABO Tables · ABO Tables | FID: 31.18 -> 27.06 | Improved |
| Use DiffRF rather than EG3D for masked completion | PhotoShape Chairs · PhotoShape Chairs masked completion (Avg) | mPSNR: 24.82 -> 27.53 | Improved |
| Use DiffRF rather than EG3D for masked completion | PhotoShape Chairs · PhotoShape Chairs masked completion (Avg) | FID: 30.67 -> 27.78 | Improved |
Later papers that compared with this method
7 later accepted papers report DiffRF as a baseline in their results tables; 7 of them report a better number on most of the results they share with it.
- GaussianGPT: Towards Autoregressive 3D Gaussian Scene Generation (ECCV 2026 · better on 1 of 1 shared result)
- DiffTF++: 3D-Aware Diffusion Transformer for Large-Vocabulary 3D Generation (TPAMI 2025 · better on 2 of 2 shared results)
- DIRECT-3D: Learning Direct Text-to-3D Generation on Massive Noisy 3D Data (CVPR 2024 · better on 2 of 2 shared results)
- Large-Vocabulary 3D Diffusion Model with Transformer (ICLR 2024 · better on 1 of 1 shared result)
- LN3Diff: Scalable Latent Neural Fields Diffusion for Speedy 3D Generation (ECCV 2024 · better on 1 of 1 shared result)
- GOEmbed: Gradient Origin Embeddings for Representation Agnostic 3D Feature Learning (ECCV 2024 · better on 1 of 1 shared result)
- Autodecoding Latent 3D Diffusion Models (NeurIPS 2023 · better on 4 of 4 shared results)
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
- DIFIX3D+: Improving 3D Reconstructions with Single-Step Diffusion Models (CVPR 2025 · similarity 0.90)
- Denoising Diffusion via Image-Based Rendering (ICLR 2024 · similarity 0.90)
- HIFA: High-fidelity Text-to-3D Generation with Advanced Diffusion Guidance (ICLR 2024 · similarity 0.90)
- CVT-xRF: Contrastive In-Voxel Transformer for 3D Consistent Radiance Fields from Sparse Inputs (CVPR 2024 · similarity 0.90)
- SIGNeRF: Scene Integrated Generation for Neural Radiance Fields (CVPR 2024 · similarity 0.89)
- Efficient-3Dim: Learning a Generalizable Single-image Novel-view Synthesizer in One Day (ICLR 2024 · similarity 0.89)
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.
