Invertible Neural Network Adapter for One-Step Flow Matching in Robot Manipulation

2026-06-17

AI Summary

This work addresses the challenge of balancing inference speed and action quality in high-dimensional robotic manipulation under multimodal observations. The authors propose a flow-matching-based invertible neural adapter that generates high-quality dexterous actions through single-step denoising in an invertible latent space. This approach introduces, for the first time, a flow matching framework amenable to single-step inference while effectively fusing visual, linguistic, and proprioceptive inputs. Experimental results demonstrate that the method achieves state-of-the-art or near state-of-the-art performance across multiple simulated benchmarks. On real robots, it reduces the average inference latency of vision-language-action (VLA) models from 110 ms to 61 ms without compromising task success rates, thereby significantly overcoming the efficiency limitations inherent in conventional iterative strategies.

Abstract

This paper presents an invertible neural network adapter for general robotic manipulation, designed to generate precise high-dimensional actions conditioned on multimodal observations, including visual, linguistic, and proprioceptive inputs, through a one-step denoising process. Built upon a flow-matching formulation, the proposed adapter effectively constrains the action generation trajectory within an invertible latent space, thereby enabling efficient and high-quality dexterous action synthesis with only a single inference step. Compared with conventional iterative flow-matching policies, the proposed framework substantially reduces inference complexity while maintaining strong action prediction accuracy and stability. Extensive experiments are conducted across a diverse set of simulation benchmarks and real-world robotic platforms to evaluate the effectiveness of the proposed method. Across simulation benchmarks, the proposed adapter consistently demonstrates superior or near state-of-the-art performance on a wide range of manipulation tasks. Furthermore, real-world experiments reveal a significant improvement in inference efficiency for vision-language-action (VLA) models, reducing the average inference latency from 110 ms to 61 ms while maintaining strong task performance.

Problem

• robotic manipulation
• multimodal observations
• action generation
• inference efficiency
• high-dimensional actions

Innovation

• Invertible Neural Network
• One-Step Flow Matching
• Robot Manipulation
• Multimodal Conditioning
• Efficient Inference