Multi-Modal Manipulation via Multi-Modal Policy Consensus
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911596359450624 |
|---|---|
| author | Chen, Haonan Xu, Jiaming Chen, Hongyu Hong, Kaiwen Huang, Binghao Liu, Chaoqi Mao, Jiayuan Li, Yunzhu Du, Yilun Driggs-Campbell, Katherine |
| author_facet | Chen, Haonan Xu, Jiaming Chen, Hongyu Hong, Kaiwen Huang, Binghao Liu, Chaoqi Mao, Jiayuan Li, Yunzhu Du, Yilun Driggs-Campbell, Katherine |
| contents | Effectively integrating diverse sensory modalities is crucial for robotic manipulation. However, the typical approach of feature concatenation is often suboptimal: dominant modalities such as vision can overwhelm sparse but critical signals like touch in contact-rich tasks, and monolithic architectures cannot flexibly incorporate new or missing modalities without retraining. Our method factorizes the policy into a set of diffusion models, each specialized for a single representation (e.g., vision or touch), and employs a router network that learns consensus weights to adaptively combine their contributions, enabling incremental of new representations. We evaluate our approach on simulated manipulation tasks in {RLBench}, as well as real-world tasks such as occluded object picking, in-hand spoon reorientation, and puzzle insertion, where it significantly outperforms feature-concatenation baselines on scenarios requiring multimodal reasoning. Our policy further demonstrates robustness to physical perturbations and sensor corruption. We further conduct perturbation-based importance analysis, which reveals adaptive shifts between modalities. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_23468 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Multi-Modal Manipulation via Multi-Modal Policy Consensus Chen, Haonan Xu, Jiaming Chen, Hongyu Hong, Kaiwen Huang, Binghao Liu, Chaoqi Mao, Jiayuan Li, Yunzhu Du, Yilun Driggs-Campbell, Katherine Robotics Artificial Intelligence Machine Learning Effectively integrating diverse sensory modalities is crucial for robotic manipulation. However, the typical approach of feature concatenation is often suboptimal: dominant modalities such as vision can overwhelm sparse but critical signals like touch in contact-rich tasks, and monolithic architectures cannot flexibly incorporate new or missing modalities without retraining. Our method factorizes the policy into a set of diffusion models, each specialized for a single representation (e.g., vision or touch), and employs a router network that learns consensus weights to adaptively combine their contributions, enabling incremental of new representations. We evaluate our approach on simulated manipulation tasks in {RLBench}, as well as real-world tasks such as occluded object picking, in-hand spoon reorientation, and puzzle insertion, where it significantly outperforms feature-concatenation baselines on scenarios requiring multimodal reasoning. Our policy further demonstrates robustness to physical perturbations and sensor corruption. We further conduct perturbation-based importance analysis, which reveals adaptive shifts between modalities. |
| title | Multi-Modal Manipulation via Multi-Modal Policy Consensus |
| topic | Robotics Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2509.23468 |