Research map / Video / Video generation
Drag Anything: Motion Control for Anything using Entity Representation
Weijia Wu, Zhuang Li, Yuchao Gu, Rui Zhao, Yefei He, David Junhao Zhang, Mike Zheng Shou, Yan Li, Tingting Gao, Zhang Di
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 entity representation (diffusion features of the selected entity) to conditioning (ablation) | DragAnything · VIPSeg val | ObjMC: 410.7 -> 318.4 | Improved |
| add 2D Gaussian representation to conditioning (ablation) | DragAnything · VIPSeg val | ObjMC: 410.7 -> 339.3 | Improved |
| combine entity and 2D Gaussian representations (ablation) | DragAnything · VIPSeg val | ObjMC: 318.4 -> 305.7 | Improved |
| add entity and Gaussian representations (ablation) | DragAnything · VIPSeg val | FVD: 496.3 -> 494.8 | No clear effect |
| restrict MSE loss to entity mask (ablation) | DragAnything · VIPSeg val | ObjMC: 311.1 -> 305.7 | Improved |
| replace pixel-trajectory control (DragNUWA) with entity-representation control | SVD-based drag video generation · VIPSeg val | FID: 39.8 -> 33.5 | Improved |
| replace pixel-trajectory control with entity-representation control | SVD-based drag video generation · VIPSeg val | FVD: 519.3 -> 494.8 | Improved |
Later papers that compared with this method
20 later accepted papers report DragAnything as a baseline in their results tables; 17 of them report a better number on most of the results they share with it.
- OmniDrag: Enabling Motion Control for Omnidirectional Image-to-Video Generation (IJCV 2026 · better on 7 of 7 shared results)
- ObjCtrl-2.5D: Training-free Object Control with Camera Poses (IJCV 2026 · better on 1 of 6 shared results)
- Time-to-Move: Training-Free Motion-Controlled Video Generation via Dual-Clock Denoising (ICLR 2026 · better on 2 of 4 shared results)
- ManipDreamer3D: Synthesizing Plausible Robotic Manipulation Video with Occupancy-aware 3D Trajectory (AAAI 2026 · better on 3 of 3 shared results)
- STANCE: Controllable Video Generation for Structured Dynamics via Sparse-To-dense ANChored Encoding (ECCV 2026 · better on 2 of 2 shared results)
- Learning Video Generation for Robotic Manipulation with Collaborative Trajectory Control (ICLR 2026 · better on 1 of 1 shared result)
- PoseTraj: Pose-Aware Trajectory Control in Video Diffusion (CVPR 2025 · better on 8 of 9 shared results)
- TrackGo: A Flexible and Efficient Method for Controllable Video Generation (AAAI 2025 · better on 8 of 8 shared results)
12 more not listed.
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
- LeviTor: 3D Trajectory Oriented Image-to-Video Synthesis (CVPR 2025 · similarity 0.85)
- 3DTrajMaster: Mastering 3D Trajectory for Multi-Entity Motion in Video Generation (ICLR 2025 · similarity 0.85)
- AnimateAnything: Consistent and Controllable Animation for Video Generation (CVPR 2025 · similarity 0.83)
- MotionPro: A Precise Motion Controller for Image-to-Video Generation (CVPR 2025 · similarity 0.83)
- Dragin3D: Image Editing by Dragging in 3D Space (CVPR 2025 · similarity 0.83)
- Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance (NeurIPS 2025 · similarity 0.82)
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.
