Metraipolis › archive › dc_c7ddeaa53d73

Video2Reaction: Training Foundation Video Models to Predict Audience Reaction — arXiv:2609.01816

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.01816
License (as archived)
CC0-1.0
Archived
2026-10-17T05:49:43.697Z
SHA-256 of the archived text
53de5e5ea42c16e5406d1b7d2505ed425232607a25eb24f1f8ceb5621eb6203d
Stable URL of this page
https://metraipolis.com/documento/dc_c7ddeaa53d73

How to cite

Video2Reaction: Training Foundation Video Models to Predict Audience Reaction — arXiv:2609.01816 — arXiv (en), https://arxiv.org/abs/2609.01816 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c7ddeaa53d73 (Metraipolis archive dc_c7ddeaa53d73, sha256 53de5e5ea42c16e5406d1b7d2505ed425232607a25eb24f1f8ceb5621eb6203d)
BibTeX
@misc{dcc7ddeaa53d73,
  title = {Video2Reaction: Training Foundation Video Models to Predict Audience Reaction — arXiv:2609.01816},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.01816},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c7ddeaa53d73; sha256 53de5e5ea42c16e5406d1b7d2505ed425232607a25eb24f1f8ceb5621eb6203d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.01816. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We introduce Video2Reaction, a multimodal dataset that maps short movie segments to the induced emotional reactions of viewers in the wild, as expressed through social media comments. Video2Reaction captures the natural diversity of emotional responses by aggregating reactions from online comments at scale, modeling labels as distributions over categorical emotions to better reflect the subjective and ambiguous nature of emotional perception. We benchmark two vision-language models (VLMs) finetuned with LoRA, showing that VLMs learn effectively from Video2Reaction and outperform specialized baselines on dominant reaction prediction. We further demonstrate that VLMs pre-finetuned on Video2Reaction transfer effectively to VCE, another induced emotion dataset with a different taxonomy and video domain. Notably, LLaVA-NeXT-Video-7B pre-finetuned on Video2Reaction and adapted on only 1% of VCE training data achieves a top-3 accuracy of 0.682, on par with the best reported VCE performance trained on the full dataset. The dataset is available at this https URL

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-03 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.01816 — 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 2026-10-17T05:25:58.027Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).