Recent vision-language-action (VLA) models for multi-task robot manipulation often rely on fixed camera setups and shared visual encoders, which limit their performance under occlusions and during cross-task transfer. To address these challenges, we propose Task-aware Virtual View Exploration (TVVE), a framework that learns to select task-relevant virtual camera viewpoints and dynamically re-render observations from a reconstructed scene representation using the selected viewpoints. To enable efficient view selection, we train an exploration policy in a pseudo-environment. In addition, we introduce a Task-aware Mixture-of-Experts (TaskMoE) visual encoder that routes visual features to task-specialized experts, mitigating interference in multi-task learning. To evaluate robustness under distribution shifts, we construct RLBench-OG, an out-of-distribution benchmark with visual perturbations and camera pose variations. Experiments on RLBench and RLBench-OG demonstrate that TVVE achieves higher success rates than strong baselines, while real-robot experiments further confirm its robustness to visual disturbances and unseen instructions. Code and visualizations are available at: https://hcplab-sysu.github.io/TAVP.
@article{arxiv.2508.05186,
title = {Learning to See and Act: Task-Aware Virtual View Exploration for Robotic Manipulation},
author = {Yongjie Bai and Zhouxia Wang and Yang Liu and Kaijun Luo and Yifan Wen and Mingtong Dai and Weixing Chen and Ziliang Chen and Lingbo Liu and Guanbin Li and Liang Lin},
journal= {arXiv preprint arXiv:2508.05186},
year = {2026}
}