Metraipolis › archive › dc_35f32db27cec

Density-Aware Calibration for Resource-Efficient Medical Vision-Language Models — arXiv:2611.00684

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00684
License (as archived)
CC0-1.0
Archived
2026-11-03T11:40:45.293Z
SHA-256 of the archived text
14971c067077ebc04b8ffd1558702217df3902b138e975c2c7cccccc39b8e980
Stable URL of this page
https://metraipolis.com/documento/dc_35f32db27cec

How to cite

Density-Aware Calibration for Resource-Efficient Medical Vision-Language Models — arXiv:2611.00684 — arXiv (en), https://arxiv.org/abs/2611.00684 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_35f32db27cec (Metraipolis archive dc_35f32db27cec, sha256 14971c067077ebc04b8ffd1558702217df3902b138e975c2c7cccccc39b8e980)
BibTeX
@misc{dc35f32db27cec,
  title = {Density-Aware Calibration for Resource-Efficient Medical Vision-Language Models — arXiv:2611.00684},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00684},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_35f32db27cec; sha256 14971c067077ebc04b8ffd1558702217df3902b138e975c2c7cccccc39b8e980},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00684. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Medical vision-language models (VLMs) show promise for clinical image understanding, but their computational and memory requirements hinder deployment on resource-constrained clinical hardware. Post-training quantization (PTQ) offers a practical compression strategy without retraining, yet its effectiveness depends strongly on the calibration data. We propose CoreCalib, a training-free and quantizer-agnostic framework for constructing compact calibration sets that preserve the target medical VQA distribution. CoreCalib stratifies data by clinical question class, identifies representative visual modes using frozen DINOv2 features with UMAP and HDBSCAN, and expands selected images into class-balanced question-answer pairs. On Kvasir-VQA-x1, we evaluate CoreCalib with MASQuant, GPTQ, and QIG using a fixed 128-pair budget. CoreCalib achieves the best METEOR with all three quantizers and the best ROUGE-L with MASQuant and GPTQ. With W4A8 MASQuant, it improves METEOR from 0.5672 with COCO calibration to 0.6234, recovering approximately 41% of the degradation relative to BF16. These results support CoreCalib as a data-centric component for resource-efficient medical VLM deployment. Project page: this https URL .

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/2611.00684 — 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 2026-11-03T11:06:37.441Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).