Research map / 3D vision / Gaussian splatting
FantasyStyle: Controllable Stylized Distillation for 3D Gaussian Splatting
Yitong Yang, Yinglin Wang, Changshuo Wang, Huajie Wang, Shuting He
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 Multi-View Frequency Consistency (MVFC) (ablation) | FantasyStyle · 3DGS style transfer | short-range LPIPS: 0.253 -> 0.250 | Improved |
| add Multi-View Frequency Consistency (MVFC) (ablation) | FantasyStyle · 3DGS style transfer | long-range LPIPS: 0.587 -> 0.574 | Improved |
| remove low-frequency components of multi-view latents (ablation) | 2D stylized priors · 3DGS style transfer | multi-view consistency | Improved |
| remove high-frequency components (ablation) | 2D stylized priors · 3DGS style transfer | texture quality | Hurt |
| drop reconstruction term (use controllable stylized distillation instead of SDS/DDS) (ablation) | FantasyStyle · 3DGS style transfer | brushstroke detail | Improved |
| drop reconstruction term, CFG scale (ablation) | FantasyStyle · 3DGS style transfer | CFG sensitivity | Improved |
| FantasyStyle vs StyleGaussian | FantasyStyle · 3DGS style transfer | ArtFID: 45.31 -> 43.52 | Improved |
| FantasyStyle vs SGSST | FantasyStyle · 3DGS style transfer | FID_content: 314.09 -> 261.71 | Improved |
| FantasyStyle vs StyleGaussian | FantasyStyle · 3DGS style transfer | long-range RMSE: 0.425 -> 0.462 | Hurt |
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.
- SGSST — SGSST: Scaling Gaussian Splatting Style Transfer (CVPR 2025) · better on 7 of 7 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
- Multi-StyleGS: Stylized Gaussian Splatting with Multiple Styles (AAAI 2025 · similarity 0.90)
- CLIPGaussian: Universal and Multimodal Style Transfer Based on Gaussian Splatting (NeurIPS 2025 · similarity 0.90)
- Morpheus: Text-Driven 3D Gaussian Splat Shape and Color Stylization (CVPR 2025 · similarity 0.90)
- ReGS: Reference-based Controllable Scene Stylization with Gaussian Splatting (NeurIPS 2024 · similarity 0.90)
- SGSST: Scaling Gaussian Splatting Style Transfer (CVPR 2025 · similarity 0.90)
- Styl3R: Instant 3D Stylized Reconstruction for Arbitrary Scenes and Styles (NeurIPS 2025 · 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.
