Modality-Composable Diffusion Policy via Inference-Time Distribution-level Composition
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866929761742225408 |
|---|---|
| author | Cao, Jiahang Zhang, Qiang Guo, Hanzhong Wang, Jiaxu Cheng, Hao Xu, Renjing |
| author_facet | Cao, Jiahang Zhang, Qiang Guo, Hanzhong Wang, Jiaxu Cheng, Hao Xu, Renjing |
| contents | Diffusion Policy (DP) has attracted significant attention as an effective method for policy representation due to its capacity to model multi-distribution dynamics. However, current DPs are often based on a single visual modality (e.g., RGB or point cloud), limiting their accuracy and generalization potential. Although training a generalized DP capable of handling heterogeneous multimodal data would enhance performance, it entails substantial computational and data-related costs. To address these challenges, we propose a novel policy composition method: by leveraging multiple pre-trained DPs based on individual visual modalities, we can combine their distributional scores to form a more expressive Modality-Composable Diffusion Policy (MCDP), without the need for additional training. Through extensive empirical experiments on the RoboTwin dataset, we demonstrate the potential of MCDP to improve both adaptability and performance. This exploration aims to provide valuable insights into the flexible composition of existing DPs, facilitating the development of generalizable cross-modality, cross-domain, and even cross-embodiment policies. Our code is open-sourced at https://github.com/AndyCao1125/MCDP. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_12466 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Modality-Composable Diffusion Policy via Inference-Time Distribution-level Composition Cao, Jiahang Zhang, Qiang Guo, Hanzhong Wang, Jiaxu Cheng, Hao Xu, Renjing Robotics Computer Vision and Pattern Recognition Diffusion Policy (DP) has attracted significant attention as an effective method for policy representation due to its capacity to model multi-distribution dynamics. However, current DPs are often based on a single visual modality (e.g., RGB or point cloud), limiting their accuracy and generalization potential. Although training a generalized DP capable of handling heterogeneous multimodal data would enhance performance, it entails substantial computational and data-related costs. To address these challenges, we propose a novel policy composition method: by leveraging multiple pre-trained DPs based on individual visual modalities, we can combine their distributional scores to form a more expressive Modality-Composable Diffusion Policy (MCDP), without the need for additional training. Through extensive empirical experiments on the RoboTwin dataset, we demonstrate the potential of MCDP to improve both adaptability and performance. This exploration aims to provide valuable insights into the flexible composition of existing DPs, facilitating the development of generalizable cross-modality, cross-domain, and even cross-embodiment policies. Our code is open-sourced at https://github.com/AndyCao1125/MCDP. |
| title | Modality-Composable Diffusion Policy via Inference-Time Distribution-level Composition |
| topic | Robotics Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.12466 |