Unified Condition-Action Modeling for Accurate One-Step Action Generation

2026-08-17

AI Summary

This study addresses the slow inference and low accuracy in diffusion and flow policies for robotic manipulation caused by improper conditioning. We propose UCA-Flow, a framework that unifies condition-action token spaces and dynamically reconstructs condition representations, integrated with an improved dual-channel u-v supervision mechanism to enable precise single-step action generation. Experimental results demonstrate that UCA-Flow achieves a 9.3% improvement in success rate and accelerates inference speed by 45.6 times compared to DP3. Furthermore, it significantly outperforms existing single-step flow policy baselines, effectively resolving critical bottlenecks in both efficiency and performance associated with condition-aided modeling in robot learning.

Abstract

Robot manipulation requires policies that are both accurate and efficient, as robot control must respond to changing observations under tight latency constraints. Recent diffusion and flow policies are promising, but they often treat conditions as auxiliary signals rather than jointly evolving them with action trajectories. We find that this limitation can be effectively mitigated by a \textbf{simple yet effective unified condition-action modeling design} that represents conditions and actions in a shared token space, allowing a compact model to achieve high performance while improving both inference speed and accuracy. Therefore, we propose UCA-Flow, a unified condition-action modeling framework for accurate one-step action generation. Our method unifies observation conditions, timestep conditions, interval conditions, and action tokens into a single sequence, and processes them with a Unified Condition-Action Transformer for joint condition-action representation learning. As a result, condition representations are dynamically reconstructed according to the current generation stage, highlighting information most relevant for action refinement. Furthermore, we introduce an improved dual-pass supervision scheme over $u$ and $v$ for stronger optimization of unified condition-action modeling. UCA-Flow improves the average success rate by 9.3 percentage points over the strongest baseline, while achieving $45.6\times$ and $33.4\times$ speedups over DP3 and Simple DP3, and remaining $4.3\times$ and $2.3\times$ faster than one-step FlowPolicy and MP1, respectively.

Problem

• Robot Manipulation
• Condition-Action Modeling
• One-Step Action Generation
• Diffusion Policy
• Flow Matching

Innovation

• Unified Condition-Action Modeling
• One-Step Action Generation
• Shared Token Space
• Dual-Pass Supervision
• UCA-Flow