Metraipolis › archive › dc_6be3aed7891b

One Skill Does Not Fit All: Automatic Discovery and Taxonomy-Guided Routing of Frame-Selection Skills for Long-Video Question Answering — arXiv:2609.12517

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.12517
License (as archived)
CC0-1.0
Archived
2026-10-16T03:57:38.181Z
SHA-256 of the archived text
383933f4593200ed3c7dc93b347464956c5b2bc51eef9c954379bac70deb61a9
Stable URL of this page
https://metraipolis.com/documento/dc_6be3aed7891b

How to cite

One Skill Does Not Fit All: Automatic Discovery and Taxonomy-Guided Routing of Frame-Selection Skills for Long-Video Question Answering — arXiv:2609.12517 — arXiv (en), https://arxiv.org/abs/2609.12517 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6be3aed7891b (Metraipolis archive dc_6be3aed7891b, sha256 383933f4593200ed3c7dc93b347464956c5b2bc51eef9c954379bac70deb61a9)
BibTeX
@misc{dc6be3aed7891b,
  title = {One Skill Does Not Fit All: Automatic Discovery and Taxonomy-Guided Routing of Frame-Selection Skills for Long-Video Question Answering — arXiv:2609.12517},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.12517},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6be3aed7891b; sha256 383933f4593200ed3c7dc93b347464956c5b2bc51eef9c954379bac70deb61a9},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.12517. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Long-Video Question Answering (LVQA) requires locating decisive evidence in hour-scale videos under a limited frame budget. Most training-free methods apply the same frame-selection strategy to all questions, despite substantial variation in the evidence required by different question types. Our analysis shows that the relative effectiveness of frame-selection strategies varies across semantic categories and benchmarks, motivating adaptive evidence acquisition. In this paper, we introduce AutoSkill, a source-supervised framework for automatically discovering and routing executable frame-selection skills. Starting from a small labelled source pool, LLM agents iteratively propose, implement, evaluate, and refine candidate skills. For a target benchmark, AutoSkill uses only unlabelled question and option text to induce a shared semantic taxonomy, rewrite labelled source examples into the target style, and estimate a category-to-skill mapping. Neither target videos nor target answers are used in this process. At inference time, each question is assigned one skill, which selects the frames used in a single inference of the frozen video MLLM. Across five long-video benchmark splits, AutoSkill improves Qwen2.5-VL-7B and Qwen3.5-4B by 2.4% and 1.2%, respectively, demonstrating the effectiveness of our AutoSkill.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-14 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.12517 — 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 2026-10-16T03:44:26.262Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).