@inproceedings{fu-etal-2026-videostir,
    title = "{V}ideo{S}tir: Understanding Long Videos via Spatio-Temporally Structured and Intent-Aware {RAG}",
    author = "Fu, Honghao  and
      Xu, Miao  and
      Wang, Yiwei  and
      Zhang, Dailing  and
      Liu, Jun  and
      Cai, Yujun",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.1656/",
    pages = "35793--35805",
    ISBN = "979-8-89176-390-6",
    abstract = "Scaling multimodal large language models (MLLMs) to long videos is constrained by limited context windows. While retrieval-augmented generation (RAG) is a promising remedy by organizing query-relevant visual evidence into a compact context, most existing methods (i) flatten videos into independent segments, breaking their inherent spatio-temporal structure, and (ii) depend on explicit semantic matching, which can miss cues that are implicitly relevant to the query{'}s intent. To overcome these limitations, we propose VideoStir, a structured and intent-aware long-video RAG framework. It firstly structures a video as a spatio-temporal graph at clip level, and then performs multi-hop retrieval to aggregate evidence across distant yet contextually related events. Furthermore, it introduces an MLLM-backed intent-relevance scorer that retrieves frames based on their alignment with the query{'}s reasoning intent. To support this capability, we curate IR-600K, a large-scale dataset tailored for learning frame{--}query intent alignment. Experiments show that VideoStir is competitive with state-of-the-art baselines without relying on auxiliary information, highlighting the promise of shifting long-video RAG from flattened semantic matching to structured, intent-aware reasoning. Codes and checkpoints are available at https://github.com/RomGai/VideoStir."
}