Research map / Video / Video segmentation and tracking
Disentangling Propagation and Generation for Video Prediction
Hang Gao, Huazhe Xu, Qi-Zhi Cai, Ruth Wang, Fisher Yu, Trevor Darrell
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 computed sparse occlusion map and inpainting over warp-only (ablation) | DPG · KITTI Flow | PSNR: 21.6 -> 22.3 | Improved |
| add computed sparse occlusion map and inpainting over warp-only (ablation) | DPG · KITTI Flow | LPIPS: 0.139 -> 0.114 | Improved |
| add occlusion map guidance to an auto-encoder refinement (ablation) | DPG · KITTI Flow | PSNR: 21.3 -> 22.3 | Improved |
| add occlusion map guidance to an auto-encoder refinement (ablation) | DPG · KITTI Flow | LPIPS: 0.120 -> 0.114 | Improved |
| use computed occlusion map instead of learned occlusion map (ablation) | DPG · KITTI Flow | PSNR: 21.8 -> 22.3 | Improved |
| use computed occlusion map instead of learned occlusion map (ablation) | DPG · KITTI Flow | LPIPS: 0.179 -> 0.114 | Improved |
| use computed occlusion map instead of learned map for localizing occlusions (ablation) | occlusion map · RoamingImages | IoU: 0.0411 -> 24.91 | Improved |
| use masking weight beta=10 instead of beta=1 (ablation) | DPG · KITTI Flow | PSNR: 21.2 -> 22.3 | Improved |
| use masking weight beta=10 instead of beta=0 (unmasked only) (ablation) | DPG · KITTI Flow | PSNR: 20.7 -> 22.3 | Improved |
| use beta=10 instead of beta=infinity (masked pixels only) (ablation) | DPG · KITTI Flow | PSNR: 18.4 -> 22.3 | Improved |
| disentangled flow plus inpainting over pixel-based DVF warping model | DPG vs DVF · KITTI Flow | LPIPS: 16.3 -> 11.4 | Improved |
| disentangled DPG versus ContextVP on CalTech | DPG vs ContextVP · CalTech Pedestrian | PSNR: 28.7 -> 28.2 | Mixed |
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 1 of 1 shared result
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
- Motion Consistency Model: Accelerating Video Diffusion with Disentangled Motion-Appearance Distillation (NeurIPS 2024 · similarity 0.87)
- FlowMo: Variance-Based Flow Guidance for Coherent Motion in Video Generation (NeurIPS 2025 · similarity 0.87)
- HiCoM: Hierarchical Coherent Motion for Dynamic Streamable Scenes with 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.86)
- Emergent Temporal Correspondences from Video Diffusion Transformers (NeurIPS 2025 · similarity 0.86)
- CanvasMAR: Improving Masked Autoregressive Video Prediction With Canvas (NeurIPS 2026 · similarity 0.86)
- OCAI: Improving Optical Flow Estimation by Occlusion and Consistency Aware Interpolation (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.
