Metraipolis › archive › dc_1962284cf5ba
CertVLA: Certified Defense against Physical Visual Attacks for Vision-Language-Action Models — arXiv:2608.20791
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.20791
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T10:52:58.087Z
- SHA-256 of the archived text
157a346f68083223a7c4f59a2e5bdd8af36ab7b777cc1f655d6d23863de05db4- Stable URL of this page
https://metraipolis.com/documento/dc_1962284cf5ba
How to cite
CertVLA: Certified Defense against Physical Visual Attacks for Vision-Language-Action Models — arXiv:2608.20791 — arXiv (en), https://arxiv.org/abs/2608.20791 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_1962284cf5ba (Metraipolis archive dc_1962284cf5ba, sha256 157a346f68083223a7c4f59a2e5bdd8af36ab7b777cc1f655d6d23863de05db4)
BibTeX
@misc{dc1962284cf5ba,
title = {CertVLA: Certified Defense against Physical Visual Attacks for Vision-Language-Action Models — arXiv:2608.20791},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.20791},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_1962284cf5ba; sha256 157a346f68083223a7c4f59a2e5bdd8af36ab7b777cc1f655d6d23863de05db4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.20791. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Vision-Language-Action (VLA) policies are vulnerable to localized physical perturbations, yet existing certified patch defenses target discrete labels and cannot directly certify continuous, temporally correlated actions. We introduce CertVLA, a certified defense for closed-loop VLA control under bounded patch and texture attacks. CertVLA proposes a calibrated region of behaviorally consistent actions, while deterministic covering masks ensure that at least one checked prediction is attack-free. Specifically, CertVLA normalizes action disagreement by the benign variation of each mask pair and accepts a single-mask anchor only when it remains consistent under every second mask. It then calibrates the resulting max-min-max episode score to provide finite-sample clean coverage. Conjoining query-level decisions extends the action certificate to the complete closed-loop rollout. Furthermore, we prove that against any adaptive attacker satisfying the bounded-support threat model, every rollout certified by CertVLA executes only action chunks consistent with attack-erased clean predictions. Under dual-mask rollout correctness, this consistency certificate further guarantees task success. The certificate is independent of patch content, generation method, and physical transformation. Experiments in simulation and the real world demonstrate the empirical and certified effectiveness of CertVLA against patch attacks, with additional simulation validation on texture attacks. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-24 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.20791 — 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 2026-10-18T10:48:13.566Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-08-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).