CDP: Towards Robust Autoregressive Visuomotor Policy Learning via Causal Diffusion

Jiahua Ma, Yiran Qin, Yixiong Li, Xuanqi Liao, Yulan Guo, Ruimao Zhang
Proceedings of The 9th Conference on Robot Learning, PMLR 305:4190-4205, 2025.

Abstract

Diffusion Policy (DP) enables robots to learn complex behaviors by imitating expert demonstrations through action diffusion. However, in practical applications, hardware limitations often degrade data quality, while real-time constraints restrict model inference to instantaneous state and scene observations. These limitations seriously reduce the efficacy of learning from expert demonstrations, resulting in failures in object localization, grasp planning, and long-horizon task execution. To address these challenges, we propose Causal Diffusion Policy (CDP), a novel transformer-based diffusion model that enhances action prediction by conditioning on historical action sequences, thereby enabling more coherent and context-aware visuomotor policy learning. In practice, to further mitigate the computational cost associated with autoregressive inference, a caching mechanism is also introduced to store attention key-value pairs from previous timesteps, substantially reducing redundant computations during execution. Extensive experiments in both simulated and real-world environments, spanning diverse 2D and 3D manipulation tasks, demonstrate that CDP uniquely leverages historical action sequences to achieve significantly higher accuracy than existing methods. Moreover, even when faced with degraded observation quality, CDP maintains remarkable precision by reasoning through temporal continuity, which highlights its practical robustness for robotic control under realistic, imperfect conditions.

Cite this Paper

BibTeX
@InProceedings{pmlr-v305-ma25c, title = {CDP: Towards Robust Autoregressive Visuomotor Policy Learning via Causal Diffusion}, author = {Ma, Jiahua and Qin, Yiran and Li, Yixiong and Liao, Xuanqi and Guo, Yulan and Zhang, Ruimao}, booktitle = {Proceedings of The 9th Conference on Robot Learning}, pages = {4190--4205}, year = {2025}, editor = {Lim, Joseph and Song, Shuran and Park, Hae-Won}, volume = {305}, series = {Proceedings of Machine Learning Research}, month = {27--30 Sep}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v305/main/assets/ma25c/ma25c.pdf}, url = {https://proceedings.mlr.press/v305/ma25c.html}, abstract = {Diffusion Policy (DP) enables robots to learn complex behaviors by imitating expert demonstrations through action diffusion. However, in practical applications, hardware limitations often degrade data quality, while real-time constraints restrict model inference to instantaneous state and scene observations. These limitations seriously reduce the efficacy of learning from expert demonstrations, resulting in failures in object localization, grasp planning, and long-horizon task execution. To address these challenges, we propose Causal Diffusion Policy (CDP), a novel transformer-based diffusion model that enhances action prediction by conditioning on historical action sequences, thereby enabling more coherent and context-aware visuomotor policy learning. In practice, to further mitigate the computational cost associated with autoregressive inference, a caching mechanism is also introduced to store attention key-value pairs from previous timesteps, substantially reducing redundant computations during execution. Extensive experiments in both simulated and real-world environments, spanning diverse 2D and 3D manipulation tasks, demonstrate that CDP uniquely leverages historical action sequences to achieve significantly higher accuracy than existing methods. Moreover, even when faced with degraded observation quality, CDP maintains remarkable precision by reasoning through temporal continuity, which highlights its practical robustness for robotic control under realistic, imperfect conditions.} }
Endnote
%0 Conference Paper %T CDP: Towards Robust Autoregressive Visuomotor Policy Learning via Causal Diffusion %A Jiahua Ma %A Yiran Qin %A Yixiong Li %A Xuanqi Liao %A Yulan Guo %A Ruimao Zhang %B Proceedings of The 9th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2025 %E Joseph Lim %E Shuran Song %E Hae-Won Park %F pmlr-v305-ma25c %I PMLR %P 4190--4205 %U https://proceedings.mlr.press/v305/ma25c.html %V 305 %X Diffusion Policy (DP) enables robots to learn complex behaviors by imitating expert demonstrations through action diffusion. However, in practical applications, hardware limitations often degrade data quality, while real-time constraints restrict model inference to instantaneous state and scene observations. These limitations seriously reduce the efficacy of learning from expert demonstrations, resulting in failures in object localization, grasp planning, and long-horizon task execution. To address these challenges, we propose Causal Diffusion Policy (CDP), a novel transformer-based diffusion model that enhances action prediction by conditioning on historical action sequences, thereby enabling more coherent and context-aware visuomotor policy learning. In practice, to further mitigate the computational cost associated with autoregressive inference, a caching mechanism is also introduced to store attention key-value pairs from previous timesteps, substantially reducing redundant computations during execution. Extensive experiments in both simulated and real-world environments, spanning diverse 2D and 3D manipulation tasks, demonstrate that CDP uniquely leverages historical action sequences to achieve significantly higher accuracy than existing methods. Moreover, even when faced with degraded observation quality, CDP maintains remarkable precision by reasoning through temporal continuity, which highlights its practical robustness for robotic control under realistic, imperfect conditions.
APA
Ma, J., Qin, Y., Li, Y., Liao, X., Guo, Y. & Zhang, R.. (2025). CDP: Towards Robust Autoregressive Visuomotor Policy Learning via Causal Diffusion. Proceedings of The 9th Conference on Robot Learning, in Proceedings of Machine Learning Research 305:4190-4205 Available from https://proceedings.mlr.press/v305/ma25c.html.

Related Material