Research map / 3D vision / 3D generation
DragAPart: Learning a Part-Level Motion Prior for Articulated Objects
Ruining Li, Chuanxia Zheng, Christian Rupprecht, Andrea Vedaldi
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 multi-resolution drag encoding in every LDM block instead of conv input only (ablation) | DragAPart (SD) · I.D. Drag-a-Move | PSNR: 19.97 -> 21.38 | Improved |
| inject drag conditioning at every block rather than input only (conv encoder) (ablation) | DragAPart (SD) · I.D. Drag-a-Move | PSNR: 19.97 -> 21.10 | Improved |
| replace conv flow encoder with multi-resolution drag encoding (ablation) | DragAPart (SD) · I.D. Drag-a-Move | PSNR: 21.10 -> 21.38 | Improved |
| replace conv flow encoder with multi-resolution drag encoding (ablation) | DragAPart (DiT) · I.D. Drag-a-Move | PSNR: 20.58 -> 21.11 | Improved |
| add random-texture renderings (domain randomization) to training (ablation) | DragAPart (SD) · O.O.D. Drag-a-Move | PSNR: 18.03 -> 19.74 | Improved |
| add random-texture renderings (domain randomization) to training (ablation) | DragAPart (SD) · I.D. Drag-a-Move | PSNR: 21.38 -> 21.82 | Improved |
| add random-texture renderings (domain randomization) to training (ablation) | DragAPart (SD) · I.D. Drag-a-Move | LPIPS: 0.066 -> 0.066 | No clear effect |
| use DragAPart vs prior drag-conditioned generators | DragAPart · Human3.6M | PSNR: 21.43 -> 23.82 | Improved |
| use pre-trained SD v1.5 backbone instead of DiT trained from scratch (ablation) | DragAPart · I.D. Drag-a-Move | PSNR: 21.11 -> 21.38 | Improved |
Closest papers by meaning
- Understanding Object Dynamics for Interactive Image-to-Video Synthesis (CVPR 2021 · similarity 0.87)
- Act the Part: Learning Interaction Strategies for Articulated Object Part Discovery (ICCV 2021 · similarity 0.87)
- PartRM: Modeling Part-Level Dynamics with Large Cross-State Reconstruction Model (CVPR 2025 · similarity 0.86)
- General Articulated Objects Manipulation in Real Images via Part-Aware Diffusion Process (NeurIPS 2024 · similarity 0.86)
- DIPO: Dual-State Images Controlled Articulated Object Generation Powered by Diverse Data (NeurIPS 2025 · similarity 0.86)
- NIFTY: Neural Object Interaction Fields for Guided Human Motion Synthesis (CVPR 2024 · similarity 0.86)
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.
