Metraipolis › archive › dc_94952985b613
V-STORM: A Structure-Aware Grounded Chain of Thought Dataset for Faithful and Fine-Grained Spatio-Temporal Video Reasoning — arXiv:2611.09511
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.09511
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-09T22:35:07.313Z
- SHA-256 of the archived text
5d4aaaa6780402556273edb0e11b5d686e21d72464073d31ba0f9d627f65d4d5- Stable URL of this page
https://metraipolis.com/documento/dc_94952985b613
How to cite
V-STORM: A Structure-Aware Grounded Chain of Thought Dataset for Faithful and Fine-Grained Spatio-Temporal Video Reasoning — arXiv:2611.09511 — arXiv (en), https://arxiv.org/abs/2611.09511 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_94952985b613 (Metraipolis archive dc_94952985b613, sha256 5d4aaaa6780402556273edb0e11b5d686e21d72464073d31ba0f9d627f65d4d5)
BibTeX
@misc{dc94952985b613,
title = {V-STORM: A Structure-Aware Grounded Chain of Thought Dataset for Faithful and Fine-Grained Spatio-Temporal Video Reasoning — arXiv:2611.09511},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.09511},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_94952985b613; sha256 5d4aaaa6780402556273edb0e11b5d686e21d72464073d31ba0f9d627f65d4d5},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.09511. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
While MLLMs can often identify what happens in a video, they still struggle with fine-grained spatio-temporal reasoning, which requires precise grounding when and where the supporting evidence appears and composing such evidence into a coherent reasoning chain. This challenge is exacerbated by the lack of scalable data construction and evaluation pipelines that provide granular, verifiable supervision linking reasoning steps to explicit spatio-temporal this http URL address this gap, we present V-STORM, a structure-aware framework that transforms videos into Spatio-Temporal Scene Graphs (STSGs) and derives Grounded Chain of Thought (GCoT) from validated evidence paths, making each reasoning step traceable to structured and verifiable spatio-temporal this http URL this framework, we construct a 260K-sample training set and a 3K-sample human-verified benchmark, both featuring GCoT annotations that link reasoning to explicit spatio-temporal this http URL reveal a clear separation between evidence localization and evidence composition: current MLLMs can often identify relevant spatio-temporal cues, yet still fail to integrate them into correct multi-step reasoning, exposing a bottleneck that answer-only evaluation largely obscures. Moreover, fine-tuning with V-STORM substantially improves both answer accuracy and grounding faithfulness, validating its effectiveness as supervision for faithful and evidence-grounded video reasoning. Our dataset, benchmark, and source code will be publicly released. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2611.09511 — 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 2026-11-09T22:30:47.336Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).