Metraipolis › archive › dc_fc8595dfa7af

GS-VLA: Plug-and-Play Viewpoint Canonicalization for Frozen VLA Policies via Gaussian Splatting — arXiv:2608.19066

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.19066
License (as archived)
CC0-1.0
Archived
2026-10-18T13:45:32.656Z
SHA-256 of the archived text
39c6340eaf47a5b4addeb7dee6cd1afa5efa753d8439024372a84e91489a57ab
Stable URL of this page
https://metraipolis.com/documento/dc_fc8595dfa7af

How to cite

GS-VLA: Plug-and-Play Viewpoint Canonicalization for Frozen VLA Policies via Gaussian Splatting — arXiv:2608.19066 — arXiv (en), https://arxiv.org/abs/2608.19066 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fc8595dfa7af (Metraipolis archive dc_fc8595dfa7af, sha256 39c6340eaf47a5b4addeb7dee6cd1afa5efa753d8439024372a84e91489a57ab)
BibTeX
@misc{dcfc8595dfa7af,
  title = {GS-VLA: Plug-and-Play Viewpoint Canonicalization for Frozen VLA Policies via Gaussian Splatting — arXiv:2608.19066},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.19066},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fc8595dfa7af; sha256 39c6340eaf47a5b4addeb7dee6cd1afa5efa753d8439024372a84e91489a57ab},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.19066. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

This paper proposes a lightweight, plug-and-play framework that improves robustness to viewpoint shifts in Vision-Language-Action (VLA) policies without policy retraining. To our knowledge, this is the first approach to directly leverage 3D Gaussian-based novel-view synthesis for observation-space adaptation in VLA policies. Current VLA performance relies on the implicit assumption that training and deployment camera configurations are identical. Our experiments show that even a small displacement of the camera mount can reduce the success rate on the LIBERO benchmark from about 90% to about 10% in the worst case. Prior approaches, such as large-scale fine-tuning or generative data augmentation, are computationally expensive and risk catastrophic forgetting. To address this, viewpoint shifts are reformulated as a localized novel-view synthesis problem. Under a Locality assumption, that camera perturbations remain within a small bounded region relative to the workspace, viewpoint normalization reduces to a scene- and policy-independent disocclusion task. Our work implements this idea with a 4M-parameter 3D-Gaussian canonicalizer prepended to a frozen VLA policy. Without modifying policy weights, GS-VLA improves performance across three orthogonal axes: (1) Policy architectures, (2) Unseen task suites, and (3) Perturbation scales. These results show that a lightweight visual module can recover a large fraction of the performance lost under viewpoint shift, without policy retraining.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-20 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.19066 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T13:41:17.996Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-20?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).