Research map / 3D vision / Gaussian splatting
Gaussian Splatting with Discretized SDF for Relightable Assets
Zuo-Liang Zhu, Jian Yang, Beibei Wang
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 discretized SDF with median loss to baseline (ablation) | Ours · Glossy Blender, Teapot | PSNR / SSIM: 19.22/0.8674 -> 23.26/0.9223 | Improved |
| add discretized SDF with median loss to baseline (ablation) | Ours · Glossy Blender, Angel | PSNR / SSIM: 20.23/0.8533 -> 21.09/0.8815 | Improved |
| add spherical initialization to baseline (ablation) | Ours · Glossy Blender, Teapot | PSNR / SSIM: 19.22/0.8674 -> 22.97/0.9176 | Improved |
| add spherical initialization on top of SDF (ablation) | Ours · Glossy Blender, Horse | PSNR / SSIM: 22.13/0.9120 -> 22.87/0.9284 | Improved |
| add projection-based consistency loss on top of SDF (ablation) | Ours · Glossy Blender, Horse | PSNR / SSIM: 22.13/0.9120 -> 23.21/0.9395 | Improved |
| add projection loss on top of SDF + spherical init (ablation) | Ours · Glossy Blender, Horse | PSNR / SSIM: 22.87/0.9284 -> 24.01/0.9481 | Improved |
| add projection loss on top of SDF + spherical init (ablation) | Ours · Glossy Blender, Angel | PSNR / SSIM: 21.59/0.8910 -> 22.03/0.8919 | Improved |
| add spherical init on top of SDF (ablation) | Ours · Glossy Blender, Teapot | PSNR / SSIM: 23.26/0.9223 -> 23.45/0.9222 | No clear effect |
| use discretized SDF in Gaussians instead of extra SDF network (vs GS-ROR) | GS-ROR · Glossy Blender | PSNR: 23.39 -> 24.52 | Improved |
| use discretized SDF instead of extra SDF network (memory) | GS-ROR · Glossy Blender | Memory: 22G -> 4G | Improved |
| use discretized SDF instead of GS-ROR | GS-ROR · Glossy Blender | LPIPS: 0.0769 -> 0.0762 | No clear effect |
| use discretized SDF instead of GS-ROR | GS-ROR · Shiny Blender | MAE: 7.23 -> 6.48 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- GS-IR — GS-IR: 3D Gaussian Splatting for Inverse Rendering (CVPR 2024) · 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
- Neural Signed Distance Function Inference through Splatting 3D Gaussians Pulled on Zero-Level Set (NeurIPS 2024 · similarity 0.92)
- Structure Consistent Gaussian Splatting with Matching Prior for Few-shot Novel View Synthesis (NeurIPS 2024 · similarity 0.91)
- GaussianUDF: Inferring Unsigned Distance Functions through 3D Gaussian Splatting (CVPR 2025 · similarity 0.90)
- GSDF: 3DGS Meets SDF for Improved Neural Rendering and Reconstruction (NeurIPS 2024 · similarity 0.90)
- GaussianPro: 3D Gaussian Splatting with Progressive Propagation (ICML 2024 · similarity 0.90)
- Optimized Minimal 3D Gaussian Splatting (NeurIPS 2025 · similarity 0.90)
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.
