Research map / Image generation / Autoregressive and latent generation
M 2 Diffuser: Diffusion-based Trajectory Optimization for Mobile Manipulation in 3D Scenes
Sixu Yan, Zeyu Zhang, Muzhi Han, Zaijin Wang, Qi Xie, Zhitian Li, Zhehan Li, Hangxin Liu, Xinggang Wang, Song Chun Zhu
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 trajectory optimization (guided sampling) to the diffusion planner (ablation) | M2Diffuser · Test Set One, grasping | success rate (%): 21.95 -> 30.54 | Improved |
| add trajectory optimization to the diffusion planner (ablation) | M2Diffuser · Test Set One, placement | success rate (%): 4.67 -> 22.89 | Improved |
| add trajectory optimization to the diffusion planner (ablation) | M2Diffuser · Test Set One, goal-reaching | success rate (%): 25.49 -> 30.49 | Improved |
| add trajectory optimization to the diffusion planner (ablation) | M2Diffuser · Test Set Two, grasping | success rate (%): 9.25 -> 14.33 | Improved |
| add trajectory optimization to the diffusion planner (ablation) | M2Diffuser · Test Set Two, placement | success rate (%): 5.70 -> 12.25 | Improved |
| add trajectory optimization to the diffusion planner (ablation) | M2Diffuser · Test Set Two, goal-reaching | success rate (%): 9.19 -> 12.61 | Improved |
| add trajectory optimization to the diffusion planner (ablation) | M2Diffuser · Test Set One, grasping | collision rate (%): 34.27 -> 20.29 | Improved |
| add trajectory optimization to the diffusion planner (ablation) | M2Diffuser · Test Set One, placement | collision rate (%): 42.44 -> 24.78 | Improved |
| add trajectory optimization to the diffusion planner (ablation) | M2Diffuser · Test Set Two, grasping | collision rate (%): 53.13 -> 36.42 | Improved |
| add trajectory optimization (increases solving time) (ablation) | M2Diffuser · Test Set One, grasping | solving time (s): 0.47 -> 4.74 | Hurt |
| add learned diffusion prior to guided trajectory optimization (ablation) | M2Diffuser · Goal-reaching (100 tasks) | success rate (%): 0.00 -> 51.00 | Improved |
| add learned diffusion prior to guided trajectory optimization (ablation) | M2Diffuser · Goal-reaching (100 tasks) | collision depth (m): 6.65 -> 4.79 | Improved |
Closest papers by meaning
- AC-DiT: Adaptive Coordination Diffusion Transformer for Mobile Manipulation (NeurIPS 2025 · similarity 0.90)
- Two-Steps Diffusion Policy for Robotic Manipulation via Genetic Denoising (NeurIPS 2025 · similarity 0.89)
- DynScene: Scalable Generation of Dynamic Robotic Manipulation Scenes for Embodied AI (CVPR 2025 · similarity 0.89)
- Generative Trajectory Stitching through Diffusion Composition (NeurIPS 2025 · similarity 0.89)
- GravMAD: Grounded Spatial Value Maps Guided Action Diffusion for Generalized 3D Manipulation (ICLR 2025 · similarity 0.89)
- MoManipVLA: Transferring Vision-language-action Models for General Mobile Manipulation (CVPR 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.
