Research map / Video / Video generation
MaskControl: Spatio-Temporal Control for Masked Motion Synthesis
Ekkasit Pinyoanuntapong, Muhammad Saleem, Korrawe Karunratanakul, Pu Wang, Hongfei Xue, Chen Chen, Chuan Guo, Junli Cao, Jian Ren, Sergey Tulyakov
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 Logits Regularizer (ablation) | MaskControl · HumanML3D (pelvis ablation) | FID: 0.142 -> 0.061 | Improved |
| add Logits Optimization (ablation) | MaskControl · HumanML3D (pelvis ablation) | Average Error (cm): 40.41 -> 0.98 | Improved |
| add Logits Optimization (ablation) | MaskControl · HumanML3D (pelvis ablation) | FID: 0.128 -> 0.061 | Improved |
| add Logits Regularizer (ablation) | MaskControl · HumanML3D (pelvis ablation) | Average Error (cm): 2.18 -> 0.98 | Improved |
| add joint control to the masked motion model (ablation) | MaskControl · HumanML3D (pelvis ablation) | FID: 0.095 -> 0.061 | Improved |
| add joint control to the masked motion model (ablation) | MaskControl · HumanML3D (pelvis ablation) | Foot Skating Ratio: 0.0527 -> 0.0547 | Hurt |
| increase control density from 1 frame to 100% of frames (ablation) | MaskControl · HumanML3D (pelvis ablation) | FID: 0.077 -> 0.054 | Improved |
| increase control density from 1 frame to 100% of frames (ablation) | MaskControl · HumanML3D (pelvis ablation) | Average Error (cm): 0.10 -> 1.64 | Hurt |
| masked motion model with control instead of diffusion TLControl | MaskControl · HumanML3D (train on pelvis only) | FID: 0.271 -> 0.061 | Improved |
| MaskControl vs OmniControl on multi-joint upper body editing | MaskControl · HumanML3D upper body editing | FID: 1.213 -> 0.074 | Improved |
| MaskControl vs ProgMoGen on zero-shot head height constraint | MaskControl · HSI-1 head height | Constraint Error: 0.012 -> 0.000 | Improved |
Closest papers by meaning
- MMM: Generative Masked Motion Model (CVPR 2024 · similarity 0.89)
- MotionPro: A Precise Motion Controller for Image-to-Video Generation (CVPR 2025 · similarity 0.89)
- Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance (NeurIPS 2025 · similarity 0.88)
- Enhanced Fine-Grained Motion Diffusion for Text-Driven Human Motion Synthesis (AAAI 2024 · similarity 0.88)
- TokenMotion: Decoupled Motion Control via Token Disentanglement for Human-centric Video Generation (CVPR 2025 · similarity 0.88)
- MotionDreamer: One-to-Many Motion Synthesis with Localized Generative Masked Transformer (ICLR 2025 · similarity 0.88)
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.
