Research map / 3D vision / Gaussian splatting
Stylos: Multi-View 3D Stylization with Single-Forward Gaussian Splatting
Hanzhou Liu, Jia Huang, Mi Lu, Srikanth Saripalli, Peng Jiang
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 Global CrossBlock instead of Global+Frame hybrid (ablation) | Stylos · CO3D pizza | PSNR: 19.78 -> 20.57 | Improved |
| use Frame CrossBlock only instead of hybrid (ablation) | Stylos · CO3D pizza | PSNR: 19.78 -> 19.72 | No clear effect |
| use Global CrossBlock instead of Frame only (ablation) | Stylos · CO3D skateboard | PSNR: 20.93 -> 21.68 | Improved |
| use scene-level style loss instead of image-level (ablation) | Stylos · CO3D held-out scenes | ArtScore: 4.78 -> 9.12 | Improved |
| use voxel-level 3D style loss instead of scene-level (ablation) | Stylos · CO3D held-out scenes | short-range RMSE: 0.036 -> 0.034 | Improved |
| use voxel-level 3D style loss instead of image-level (ablation) | Stylos · CO3D held-out scenes | long-range LPIPS: 0.157 -> 0.153 | Improved |
| use voxel-level 3D style loss instead of scene-level (ablation) | Stylos · CO3D held-out scenes | ArtScore: 9.12 -> 9.15 | No clear effect |
| use voxel-level 3D style loss instead of image-level (ablation) | Stylos · CO3D held-out scenes | long-range RMSE: 0.142 -> 0.142 | No clear effect |
| increase views per batch beyond 32 (up to 64) (ablation) | Stylos · Tanks & Temples Lighthouse | visual quality | Hurt |
| use a single view per batch (ablation) | Stylos · Tanks & Temples Lighthouse | visual quality | Hurt |
| use single-forward Stylos instead of per-scene methods | Stylos · Tanks & Temples | stylization time: 0.05 s vs 14.7 m (G-Style) | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Styl3R — Styl3R: Instant 3D Stylized Reconstruction for Arbitrary Scenes and Styles (NeurIPS 2025) · better on 6 of 6 shared results
- SGSST — SGSST: Scaling Gaussian Splatting Style Transfer (CVPR 2025) · better on 2 of 2 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
- Styl3R: Instant 3D Stylized Reconstruction for Arbitrary Scenes and Styles (NeurIPS 2025 · similarity 0.93)
- Multi-StyleGS: Stylized Gaussian Splatting with Multiple Styles (AAAI 2025 · similarity 0.91)
- Morpheus: Text-Driven 3D Gaussian Splat Shape and Color Stylization (CVPR 2025 · similarity 0.91)
- ReGS: Reference-based Controllable Scene Stylization with Gaussian Splatting (NeurIPS 2024 · similarity 0.91)
- CLIPGaussian: Universal and Multimodal Style Transfer Based on Gaussian Splatting (NeurIPS 2025 · similarity 0.90)
- FPRF: Feed-Forward Photorealistic Style Transfer of Large-Scale 3D Neural Radiance Fields (AAAI 2024 · 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.
