Metraipolis › archive › dc_9596f60d6fa6

Target-Checked Reliability Score Refinement for Video Question Answering — arXiv:2609.13288

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.13288
License (as archived)
CC0-1.0
Archived
2026-10-16T01:52:37.528Z
SHA-256 of the archived text
436131cde101168fe35e30b4ad6818119bcf10c528eeb44a7a4d5785045a17d1
Stable URL of this page
https://metraipolis.com/documento/dc_9596f60d6fa6

How to cite

Target-Checked Reliability Score Refinement for Video Question Answering — arXiv:2609.13288 — arXiv (en), https://arxiv.org/abs/2609.13288 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9596f60d6fa6 (Metraipolis archive dc_9596f60d6fa6, sha256 436131cde101168fe35e30b4ad6818119bcf10c528eeb44a7a4d5785045a17d1)
BibTeX
@misc{dc9596f60d6fa6,
  title = {Target-Checked Reliability Score Refinement for Video Question Answering — arXiv:2609.13288},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.13288},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9596f60d6fa6; sha256 436131cde101168fe35e30b4ad6818119bcf10c528eeb44a7a4d5785045a17d1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.13288. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Video-language models can answer multiple-choice questions with high confidence yet be wrong. We study whether answer-level reliability scores can be improved under target shift without retraining the models or changing their answers. We collect option-probability lists from three fixed video-language models under four deterministic video samplings and represent cross-view changes and cross-model agreement as a response graph. Using a labeled target pilot, we compare the original score, defined as the probability assigned to the chosen answer, with a histogram-based gradient-boosting (HGB) score trained on the development datasets and a regularized logistic-regression score trained on the target pilot. A candidate replaces the original score only when repeated video-level checks indicate a positive, stable improvement. We develop this rule on public VideoQA benchmarks and Video Hallucination Diagnosis (VHD), a controlled diagnostic dataset for shared high-confidence errors. Ranking quality is measured by the area under the risk-coverage curve (AURC), where lower is better. On a held-out 963-question HERBench split, the method reduces mean AURC across the three models by 16.64% (95% confidence interval (CI), 12.12 to 22.61%); the smallest model-level gain is 11.39%. On a separate held-out 911-question Perception Test split, the mean reduction is 18.87% (95% CI, 15.43 to 22.14%). For InternVL3.5, the target check retains the original scores. Using the same outputs, the method outperforms seven training-free baselines in mean AURC on both datasets. It also improves AUROC, reduces calibration error, and lowers the error rate at 50% coverage by 6.50 and 6.58 percentage points.

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/2609.13288 — 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 2026-10-16T01:25:23.497Z 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).