Metraipolis › archive › dc_e5dbbd7a5bf8

CinematicVQA: Benchmarking Film-Grammar Reasoning in Large Vision-Language Models — arXiv:2609.28813

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.28813
License (as archived)
CC0-1.0
Archived
2026-10-08T06:44:14.478Z
SHA-256 of the archived text
f62449eb2b44dd9a857f9618d9766fe95104fa6cae0557d8bdbed7adeacde845
Stable URL of this page
https://metraipolis.com/documento/dc_e5dbbd7a5bf8

How to cite

CinematicVQA: Benchmarking Film-Grammar Reasoning in Large Vision-Language Models — arXiv:2609.28813 — arXiv (en), https://arxiv.org/abs/2609.28813 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e5dbbd7a5bf8 (Metraipolis archive dc_e5dbbd7a5bf8, sha256 f62449eb2b44dd9a857f9618d9766fe95104fa6cae0557d8bdbed7adeacde845)
BibTeX
@misc{dce5dbbd7a5bf8,
  title = {CinematicVQA: Benchmarking Film-Grammar Reasoning in Large Vision-Language Models — arXiv:2609.28813},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.28813},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e5dbbd7a5bf8; sha256 f62449eb2b44dd9a857f9618d9766fe95104fa6cae0557d8bdbed7adeacde845},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.28813. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Cinematography, the craft of visual storytelling through framing, lighting, and camera operation, fundamentally shapes how audiences perceive and emotionally engage with video content. While Large Vision Language Models (LVLMs) have made remarkable progress in video question answering, existing benchmarks primarily focus on identifying low-level techniques rather than understanding their storytelling impact. To address this, we introduce CinematicVQA, the first-of-its-kind benchmark for cinematic video understanding that goes beyond technique recognition to evaluate film-grammar reasoning, utilizing our introduced Cinematic Scene Graph (CSG), a structured representation that links filming techniques to their perceptual effects and narrative functions. Through comprehensive evaluation of state-of-the-art LVLMs, we reveal a striking semantic gap: models consistently perform higher on describing visual presentations than on identifying the underlying techniques. Surprisingly, Chain-of-Thought prompting fails to provide consistent gains and degrades performance for most models, suggesting that current LVLMs lack sufficient cinematic domain knowledge to benefit from step-by-step reasoning. Fine-tuning on \textsc{CinematicVQA-train} yields consistent improvements, particularly for narrative function and multi-hop reasoning. Overall, \textsc{CinematicVQA} serves both as a rigorous benchmark for cinematic evaluation in LVLMs and as a practical dataset for training more film-aware video models.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-25 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.28813 — 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 2026-10-08T06:09:29.781Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).