Look Before Acting: Enhancing Vision Foundation Representations for Vision-Language-Action Models

TL;DR

This paper introduces DeepVision-VLA, a framework that enhances visual representations in vision-language-action models for robotic manipulation by integrating multi-level visual features and pruning irrelevant tokens, leading to improved performance.

Contribution

It proposes a novel VL-MoT framework with shared attention and visual pruning techniques to improve visual grounding in VLA models, with significant performance gains.

Findings

DeepVision-VLA outperforms previous methods by 9.0% and 7.5% on simulated and real tasks.

Sensitivity to visual tokens decreases in deeper layers during action generation.

Shared attention and visual pruning improve manipulation accuracy.

Abstract

Vision-Language-Action (VLA) models have recently emerged as a promising paradigm for robotic manipulation, in which reliable action prediction critically depends on accurately interpreting and integrating visual observations conditioned on language instructions. Although recent works have sought to enhance the visual capabilities of VLA models, most approaches treat the LLM backbone as a black box, providing limited insight into how visual information is grounded into action generation. Therefore, we perform a systematic analysis of multiple VLA models across different action-generation paradigms and observe that sensitivity to visual tokens progressively decreases in deeper layers during action generation. Motivated by this observation, we propose \textbf{DeepVision-VLA}, built on a \textbf{Vision-Language Mixture-of-Transformers (VL-MoT)} framework. This framework enables shared attention between the vision foundation model and the VLA backbone, injecting multi-level visual features from the vision expert into deeper layers of the VLA backbone to enhance visual representations for precise and complex manipulation. In addition, we introduce \textbf{Action-Guided Visual Pruning (AGVP)}, which leverages shallow-layer attention to prune irrelevant visual tokens while preserving task-relevant ones, reinforcing critical visual cues for manipulation with minimal computational overhead. DeepVision-VLA outperforms prior state-of-the-art methods by 9.0\% and 7.5\% on simulated and real-world tasks, respectively, providing new insights for the design of visually enhanced VLA models.