Research map / 3D vision / 3D generation
SPECSIA: Stylization Dataset for Novel-View Enhancement in Drawing-based 3D Animation
Kyuwon Kim, Sunjae Yoon, Chang D. Yoo
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 lightweight per-character adaptation (ablation) | DraViE · Novel-view motions | CLIP: 0.838 -> 0.859 | Improved |
| add lightweight per-character adaptation (ablation) | DraViE · Novel-view motions | FID: 221.48 -> 206.13 | Improved |
| add lightweight per-character adaptation (ablation) | DraViE · Frontal-view motions | FID: 207.15 -> 201.31 | Improved |
| add positional hint Z^pos (ablation) | DraViE · Novel-view motions | FID: 318.39 -> 206.13 | Improved |
| add positional hint Z^pos (ablation) | DraViE · Novel-view motions | CLIP: 0.844 -> 0.859 | Improved |
| add positional hint Z^pos (ablation) | DraViE · Frontal-view motions | FID: 205.81 -> 201.31 | Improved |
| replace sample-wise refinement (DSU) with SPECSIA-pretrained DraViE | DrawingSpinUp pipeline · Novel-view motions | FID: 280.83 -> 206.13 | Improved |
| replace OSF post-correction with DraViE | OSF · Novel-view motions | FID: 270.21 -> 206.13 | Improved |
| replace DSU/OSF with DraViE in frontal view | DSU/OSF · Frontal-view motions | CLIP: 0.902 -> 0.905 | No clear effect |
| apply same pretrained DraViE to CRM backend | CRM projections · Cross-pipeline CRM | FID: 271/266/239 | Improved |
| apply same pretrained DraViE to InstantMesh backend | InstantMesh projections · Cross-pipeline InstantMesh | FID: 242/233/205 | Improved |
| replace DSU/OSF with DraViE on Amateur Drawings | DSU/OSF · Amateur Drawings novel-view | FID: 267.67 -> 213.40 | Improved |
Closest papers by meaning
- Styl3R: Instant 3D Stylized Reconstruction for Arbitrary Scenes and Styles (NeurIPS 2025 · similarity 0.86)
- StdGEN: Semantic-Decomposed 3D Character Generation from Single Images (CVPR 2025 · similarity 0.86)
- Neural 3D Strokes: Creating Stylized 3D Scenes with Vectorized 3D Strokes (CVPR 2024 · similarity 0.86)
- 3DToonify: Creating Your High-Fidelity 3D Stylized Avatar Easily from 2D Portrait Images (CVPR 2024 · similarity 0.85)
- ReGS: Reference-based Controllable Scene Stylization with Gaussian Splatting (NeurIPS 2024 · similarity 0.85)
- Morpheus: Text-Driven 3D Gaussian Splat Shape and Color Stylization (CVPR 2025 · similarity 0.85)
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.
