Research map / Video / Video generation
RoMan-4D: Learning Robot Arm Manipulation from 4D World Models
Kaichen Zhou, Yuzhen Chen, Fangneng Zhan, Hang Hua, Grace Chen, Xinhai Chang, Ao Qu, Yilun Du, Zhuang Liu, Paul Pu Liang, Mengyu Wang
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 geometry-feature distillation branch (geometry flow) to video DiT | GEM-4D · Droid (real) | AbsRel: 21.39 -> 20.13 vs Wan 2.2-14B | Improved |
| add geometry distillation | GEM-4D · RLBench (sim) | Chamfer: 0.1488 -> 0.0702 vs Geometry-Forcing | Improved |
| add geometry distillation (real-world manipulation via AIDS) | GEM-4D · Droid human study RAIL | success rate (%): 59 -> 87 | Improved |
| geometry-consistent rollouts versus TesserAct rollouts | GEM-4D · RLBench Reach Target | success rate (%): 2 -> 82 | Improved |
| replace geometry-feature distillation with explicit depth-velocity supervision (ablation) | GEM-4D · Droid (real) | AbsRel: 20.13 -> 20.89 | Hurt |
| use VGGT features instead of the default geometry foundation model (ablation) | GEM-4D · Droid (real) | AbsRel: 20.13 -> 21.73 | Hurt |
| use VGGT features instead of default geometry model (ablation) | GEM-4D · Droid (real) | FVD: 31.82 -> 33.68 | Hurt |
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.
- Geometry-Forcing — Geometry Forcing: Marrying Video Diffusion and 3D Representation for Consistent World Modeling (ICLR 2026) · better on 3 of 3 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
- Pre-training Auto-regressive Robotic Models with 4D Representations (ICML 2025 · similarity 0.89)
- Predictive Inverse Dynamics Models are Scalable Learners for Robotic Manipulation (ICLR 2025 · similarity 0.88)
- DynaRend: Learning 3D Dynamics via Masked Future Rendering for Robotic Manipulation (NeurIPS 2025 · similarity 0.88)
- EA-WM: Event-Aware Generative World Model with Structured Kinematic-to-Visual Action Fields (NeurIPS 2026 · similarity 0.88)
- EnerVerse: Envisioning Embodied Future Space for Robotics Manipulation (NeurIPS 2025 · similarity 0.88)
- RoboAlign-R1: Distilled Multimodal Reward Alignment for Robot Video World Models (NeurIPS 2026 · 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.
