Metraipolis › archive › dc_fd54f14630bc
RVSD: Retrieval Vision Sparse Decoding for Mitigating Visual Hallucinations in Large Vision-Language Models — arXiv:2609.02731
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.02731
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T03:38:30.946Z
- SHA-256 of the archived text
68dfa0d31bb4b22414d7d0ef94ec000bb10ee9d1dc36edd25c40e4b93f07c739- Stable URL of this page
https://metraipolis.com/documento/dc_fd54f14630bc
How to cite
RVSD: Retrieval Vision Sparse Decoding for Mitigating Visual Hallucinations in Large Vision-Language Models — arXiv:2609.02731 — arXiv (en), https://arxiv.org/abs/2609.02731 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_fd54f14630bc (Metraipolis archive dc_fd54f14630bc, sha256 68dfa0d31bb4b22414d7d0ef94ec000bb10ee9d1dc36edd25c40e4b93f07c739)
BibTeX
@misc{dcfd54f14630bc,
title = {RVSD: Retrieval Vision Sparse Decoding for Mitigating Visual Hallucinations in Large Vision-Language Models — arXiv:2609.02731},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.02731},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_fd54f14630bc; sha256 68dfa0d31bb4b22414d7d0ef94ec000bb10ee9d1dc36edd25c40e4b93f07c739},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02731. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large vision-language models have achieved remarkable success in vision-language tasks. However, they remain prone to Visual Hallucinations (VHs), undermining their reliability in real-world applications. Existing solutions typically require curated datasets, additional training, or multi-round decoding, resulting in considerable computational overhead. In this paper, we propose \textbf{RVSD} (\underline{R}etrieval \underline{V}ision \underline{S}parse \underline{D}ecoding), a training-free and plug-and-play decoding framework that, for the first time, unifies token sparsification and \textbf{Semantic-Space Visual Retrieval} (SSVR) within a single decoding pass. Within RVSD, we introduce a \textbf{semantics-directed token selection} strategy that selectively sparsifies redundant tokens while preserving critical visual information. We further propose the SSVR mechanism, which reformulates visual compensation as an on-demand cross-modal retrieval process within a shared semantic space. Extensive experiments demonstrate that RVSD achieves state-of-the-art performance in mitigating VHs while maintaining robust suppression capabilities under long-context generation settings. Our code is available here.\footnote{ this https URL }
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-03 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.02731 — 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 2026-10-17T03:13:12.566Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).