Research map / Video / Video generation
Point-to-Point Video Generation
Tsun-Hsuan Wang, Yen-Chi Cheng, Chieh Hubert Lin, Hwann-Tzong Chen, Min Sun
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 CPC loss on the prior (ablation) | SVG · SM-MNIST | SSIM S-CPC: 0.621 -> 0.729 | Improved |
| add CPC loss on the prior (ablation) | SVG · Human3.6M | MSE S-CPC: 10.83 -> 12.08 | Hurt |
| add latent-space alignment loss (ablation) | SVG + C · SM-MNIST | SSIM S-CPC: 0.729 -> 0.735 | Improved |
| add latent-space alignment loss (ablation) | SVG + C · Weizmann | SSIM S-CPC: 0.730 -> 0.767 | Improved |
| add latent-space alignment loss (ablation) | SVG + C · Human3.6M | MSE S-CPC: 12.08 -> 6.66 | Improved |
| add skip-frame training (ablation) | SVG + C + A · SM-MNIST | SSIM S-CPC: 0.735 -> 0.769 | Improved |
| add skip-frame training (ablation) | SVG + C + A · Weizmann | SSIM S-CPC: 0.767 -> 0.783 | Improved |
| add skip-frame training (ablation) | SVG + C + A · Human3.6M | MSE S-CPC: 6.66 -> 0.72 | Improved |
| add alignment loss (ablation) | SVG + C · Human3.6M | MSE S-Best: 8.25 -> 4.96 | Improved |
| full method (CPC on prior, alignment, skip-frame) versus SVG | Ours (full) · Weizmann | SSIM S-Div (1E-3): 1.992 -> 1.106 | Hurt |
| full method versus SVG | Ours (full) · SM-MNIST | SSIM S-Best: 0.780 -> 0.755 | No clear effect |
| full method versus SVG | Ours (full) · Human3.6M | MSE S-Best: 6.49 -> 4.46 | Improved |
Closest papers by meaning
- Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance (NeurIPS 2025 · similarity 0.87)
- Generative Inbetweening through Frame-wise Conditions-Driven Video Generation (CVPR 2025 · similarity 0.87)
- Motion Prompting: Controlling Video Generation with Motion Trajectories (CVPR 2025 · similarity 0.87)
- Extrapolating and Decoupling Image-to-Video Generation Models: Motion Modeling is Easier Than You Think (CVPR 2025 · similarity 0.86)
- Follow-Your-Click: Open-domain Regional Image Animation via Motion Prompts (AAAI 2025 · similarity 0.86)
- IM-Zero: Instance-level Motion Controllable Video Generation in a Zero-shot Manner (CVPR 2025 · 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.
