Metraipolis › archive › dc_aa4cd33c6e5e

MedICL: Modality-Anatomy-Aware Visual In-Context Learning for Medical Visual Question Answering — arXiv:2611.00067

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00067
License (as archived)
CC0-1.0
Archived
2026-11-03T18:52:58.164Z
SHA-256 of the archived text
39bda05d46f8c8cb5c09f477becc5a68904791d4478e129ad55c34e0ce3874e6
Stable URL of this page
https://metraipolis.com/documento/dc_aa4cd33c6e5e

How to cite

MedICL: Modality-Anatomy-Aware Visual In-Context Learning for Medical Visual Question Answering — arXiv:2611.00067 — arXiv (en), https://arxiv.org/abs/2611.00067 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_aa4cd33c6e5e (Metraipolis archive dc_aa4cd33c6e5e, sha256 39bda05d46f8c8cb5c09f477becc5a68904791d4478e129ad55c34e0ce3874e6)
BibTeX
@misc{dcaa4cd33c6e5e,
  title = {MedICL: Modality-Anatomy-Aware Visual In-Context Learning for Medical Visual Question Answering — arXiv:2611.00067},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00067},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_aa4cd33c6e5e; sha256 39bda05d46f8c8cb5c09f477becc5a68904791d4478e129ad55c34e0ce3874e6},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00067. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Training-free adaptation of medical vision-language models (LVLMs) is a deployment requirement, yet visual in-context learning (ICL) injects modality- and style-mismatched demonstrations into medical contexts, teaching a frozen model the wrong answering style. We present MedICL, which curates demonstrations instead of retraining the model. MADS selects modality-anatomy aligned demonstrations that cover complementary interpretation dimensions; W2C curates them with weak retrieval signals and a strong verification vote that checks demonstrations rather than the model's own outputs; ECC answers with W anchored demonstration sets, aggregates by abnormality-aware signatures, and re-answers once on conflict. On VQA-RAD, SLAKE, and PathVQA under the LLaVA-Med zero-shot protocol, MedICL improves the strongest zero-shot baseline by +11 - +14 points on open-ended and +7 - +14 on closed questions, recovers 71% of fine-tuned open-ended performance on VQA-RAD at zero training cost, and stays robust to shifted or corrupted demonstration banks.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2611.00067 — 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 2026-11-03T18:16:52.849Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).