Research

Paper

TESTING March 06, 2026

AnyCamVLA: Zero-Shot Camera Adaptation for Viewpoint Robust Vision-Language-Action Models

Authors

Hyeongjun Heo, Seungyeon Woo, Sang Min Kim, Junho Kim, Junho Lee, Yonghyeon Lee, Young Min Kim

Abstract

Despite remarkable progress in Vision-Language-Action models (VLAs) for robot manipulation, these large pre-trained models require fine-tuning to be deployed in specific environments. These fine-tuned models are highly sensitive to camera viewpoint changes that frequently occur in unstructured environments. In this paper, we propose a zero-shot camera adaptation framework without additional demonstration data, policy fine-tuning, or architectural modification. Our key idea is to virtually adjust test-time camera observations to match the training camera configuration in real-time. For that, we use a recent feed-forward novel view synthesis model which outputs high-quality target view images, handling both extrinsic and intrinsic parameters. This plug-and-play approach preserves the pre-trained capabilities of VLAs and applies to any RGB-based policy. Through extensive experiments on the LIBERO benchmark, our method consistently outperforms baselines that use data augmentation for policy fine-tuning or additional 3D-aware features for visual input. We further validate that our approach constantly enhances viewpoint robustness in real-world robotic manipulation scenarios, including settings with varying camera extrinsics, intrinsics, and freely moving handheld cameras.

Metadata

arXiv ID: 2603.05868
Provider: ARXIV
Primary Category: cs.RO
Published: 2026-03-06
Fetched: 2026-03-09 06:05

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.05868v1</id>\n    <title>AnyCamVLA: Zero-Shot Camera Adaptation for Viewpoint Robust Vision-Language-Action Models</title>\n    <updated>2026-03-06T03:44:23Z</updated>\n    <link href='https://arxiv.org/abs/2603.05868v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.05868v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Despite remarkable progress in Vision-Language-Action models (VLAs) for robot manipulation, these large pre-trained models require fine-tuning to be deployed in specific environments. These fine-tuned models are highly sensitive to camera viewpoint changes that frequently occur in unstructured environments. In this paper, we propose a zero-shot camera adaptation framework without additional demonstration data, policy fine-tuning, or architectural modification. Our key idea is to virtually adjust test-time camera observations to match the training camera configuration in real-time. For that, we use a recent feed-forward novel view synthesis model which outputs high-quality target view images, handling both extrinsic and intrinsic parameters. This plug-and-play approach preserves the pre-trained capabilities of VLAs and applies to any RGB-based policy. Through extensive experiments on the LIBERO benchmark, our method consistently outperforms baselines that use data augmentation for policy fine-tuning or additional 3D-aware features for visual input. We further validate that our approach constantly enhances viewpoint robustness in real-world robotic manipulation scenarios, including settings with varying camera extrinsics, intrinsics, and freely moving handheld cameras.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.RO'/>\n    <published>2026-03-06T03:44:23Z</published>\n    <arxiv:comment>Under review, Project Page: https://heo0224.github.io/AnyCamVLA/</arxiv:comment>\n    <arxiv:primary_category term='cs.RO'/>\n    <author>\n      <name>Hyeongjun Heo</name>\n    </author>\n    <author>\n      <name>Seungyeon Woo</name>\n    </author>\n    <author>\n      <name>Sang Min Kim</name>\n    </author>\n    <author>\n      <name>Junho Kim</name>\n    </author>\n    <author>\n      <name>Junho Lee</name>\n    </author>\n    <author>\n      <name>Yonghyeon Lee</name>\n    </author>\n    <author>\n      <name>Young Min Kim</name>\n    </author>\n  </entry>"
}