Research map / 3D vision / 3D generation
Stroke3D: Lifting 2D strokes into rigged 3D model via latent diffusion models
Ruisi Zhao, Haoren Zheng, Zongxin Yang, Hehe Fan, Yi Yang
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 TextuRig data (ablation) | SKDream · skeleton-to-mesh | SKA score Mean_Inst: 80.43 -> 82.37 | Improved |
| add TextuRig data (ablation) | SKDream · skeleton-to-mesh | SKA score Mean_Class: 74.38 -> 76.84 | Improved |
| add SKA-DPO (ablation) | SKDream · skeleton-to-mesh | SKA score Mean_Inst: 80.43 -> 85.57 | Improved |
| add SKA-DPO on top of TextuRig (ablation) | SKDream + TextuRig · skeleton-to-mesh | SKA score Mean_Inst: 82.37 -> 87.83 | Improved |
| add SKA-DPO on top of TextuRig (ablation) | SKDream + TextuRig · skeleton-to-mesh | SKA score Mean_Class: 76.84 -> 84.36 | Improved |
| use preference margin 0.10 instead of 0.05 (ablation) | SKA-DPO · skeleton-to-mesh | SKA score Mean_Inst: 86.95 -> 87.83 | Improved |
| use preference margin 0.10 instead of 0.20 (ablation) | SKA-DPO · skeleton-to-mesh | SKA score Mean_Inst: 86.96 -> 87.83 | Improved |
| use preference margin 0.15 instead of 0.10 (ablation) | SKA-DPO · skeleton-to-mesh | SKA score Mean_Inst: 87.83 -> 87.84 | No clear effect |
| add rotational alignment of dataset (ablation) | Sk-DiT · skeleton generation | CD-J2J (All): 0.0493 -> 0.0475 | Improved |
| add descriptive and viewpoint tags (ablation) | Sk-DiT · skeleton generation | CD-J2J (All): 0.0524 -> 0.0475 | Improved |
| use full training data instead of reduced subset (ablation) | Sk-DiT · skeleton generation | CD-J2J (All): 0.0544 -> 0.0475 | Improved |
| add 2D stroke structural condition Jxy (ablation) | Sk-DiT · overfitting on 1 or 5 samples | training loss | 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.
- MagicArticulate — MagicArticulate: Make Your 3D Models Articulation-Ready (CVPR 2025) · better on 2 of 3 shared results
- RigNet — RigNet: Repetitive Image Guided Network for Depth Completion (ECCV 2022) · better on 3 of 3 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
- Rigel3D: Rig-aware Latents for Animation-Ready 3D Asset Generation (NeurIPS 2026 · similarity 0.90)
- Puppeteer: Rig and Animate Your 3D Models (NeurIPS 2025 · similarity 0.88)
- HumanRig: Learning Automatic Rigging for Humanoid Character in a Large Scale Dataset (CVPR 2025 · similarity 0.88)
- TapMo: Shape-aware Motion Generation of Skeleton-free Characters (ICLR 2024 · similarity 0.88)
- MagicArticulate: Make Your 3D Models Articulation-Ready (CVPR 2025 · similarity 0.87)
- CraftsMan3D: High-fidelity Mesh Generation with 3D Native Diffusion and Interactive Geometry Refiner (CVPR 2025 · similarity 0.87)
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.
