Metraipolis › archive › dc_064fabefee13

Two-Stage Mixture-of-LoRA for Multi-Task Medical Vision-Language Learning — arXiv:2609.14350

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.14350
License (as archived)
CC0-1.0
Archived
2026-10-16T01:52:56.447Z
SHA-256 of the archived text
fd472b30ce61d850f85ff8c46a6c248f44e63cd4720a351d2fcff4299bfd03f2
Stable URL of this page
https://metraipolis.com/documento/dc_064fabefee13

How to cite

Two-Stage Mixture-of-LoRA for Multi-Task Medical Vision-Language Learning — arXiv:2609.14350 — arXiv (en), https://arxiv.org/abs/2609.14350 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_064fabefee13 (Metraipolis archive dc_064fabefee13, sha256 fd472b30ce61d850f85ff8c46a6c248f44e63cd4720a351d2fcff4299bfd03f2)
BibTeX
@misc{dc064fabefee13,
  title = {Two-Stage Mixture-of-LoRA for Multi-Task Medical Vision-Language Learning — arXiv:2609.14350},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.14350},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_064fabefee13; sha256 fd472b30ce61d850f85ff8c46a6c248f44e63cd4720a351d2fcff4299bfd03f2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.14350. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Medical vision-language models (VLMs) allow a single model to perform clinical image analysis tasks ranging from diagnosis classification to report generation. However, joint adaptation is challenged by heterogeneous output formats, conflicting task gradients, and imbalanced training data. Hence, we present \textbf{Two-Stage Mixture-of-LoRA}, a framework built on MedGemma-1.5-4B. The framework uses a shared-specific Mixture-of-LoRA architecture comprising one shared LoRA and six task-specific expert LoRAs, together with a two-stage training procedure. In Stage 1, we jointly train the shared LoRA and all task-specific expert LoRAs on all tasks. In Stage 2, we first freeze the backbone, the shared LoRA, and all non-target experts, and refine one task expert at a time. Classification and regression then receive an additional modality-balanced continuation, in which smaller modality groups are repeated to match the largest group. In the FLARE 2026 Task 3 test sets, the proposed method achieves 0.85 balanced accuracy for classification, 0.48 micro-F1 for multi-label classification, 0.79 detection F1, and 17.39 regression MAE. Code is available at this https URL .

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-15 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.14350 — 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 2026-10-16T01:25:23.573Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-15?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).