Research map / Video / Video generation
Learning to Forecast and Refine Residual Motion for Image-to-Video Generation
Long Zhao, Xi Peng, Yu Tian, Mubbasir Kapadia, Dimitris Metaxas
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 dense connections in the decoder of G_M (ablation) | G_M without refinement · MUG facial expression retargeting | ACD-I: 0.459 -> 0.252 | Improved |
| add dense connections in the decoder of G_M (ablation) | G_M without refinement · MUG facial expression retargeting | ACD-C: 0.155 -> 0.140 | Improved |
| add dense connections in the decoder of G_M (ablation) | G_M without refinement · Penn Action (GT motion maps) | MSE: 0.027 -> 0.014 | Improved |
| add the refinement network G_R (ablation) | G_M with dense connections · MUG facial expression retargeting | ACD-I: 0.252 -> 0.184 | Improved |
| add the refinement network G_R (ablation) | G_M with dense connections · MUG facial expression retargeting | ACD-C: 0.140 -> 0.107 | Improved |
| add the refinement network G_R (ablation) | G_M with dense connections · Penn Action (GT motion maps) | MSE: 0.014 -> 0.011 | Improved |
| use residual motion/content mask formulation instead of plain difference map (Villegas et al.) (ablation) | motion forecasting network · Penn Action | MSE: 0.030 -> 0.023 | Improved |
| use residual motion/content mask formulation instead of plain difference map (Villegas et al.), GT motion maps (ablation) | motion forecasting network · Penn Action (GT motion maps) | MSE: 0.025 -> 0.011 | Improved |
| use 3DMM-based condition with temporal refinement versus MoCoGAN | video generator · MUG facial expression retargeting | ACD-I: 0.291 -> 0.184 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MoCoGAN — MoCoGAN: Decomposing Motion and Content for Video Generation (CVPR 2018) · better on 2 of 2 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
- Snap Video: Scaled Spatiotemporal Transformers for Text-to-Video Synthesis (CVPR 2024 · similarity 0.88)
- Generative Pre-trained Autoregressive Diffusion Transformer (NeurIPS 2025 · similarity 0.87)
- Move-in-2D: 2D-Conditioned Human Motion Generation (CVPR 2025 · similarity 0.87)
- Motion Prompting: Controlling Video Generation with Motion Trajectories (CVPR 2025 · similarity 0.87)
- Generative Inbetweening: Adapting Image-to-Video Models for Keyframe Interpolation (ICLR 2025 · similarity 0.87)
- VideoJAM: Joint Appearance-Motion Representations for Enhanced Motion Generation in Video Models (ICML 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.
