@inproceedings{bhatnagar-etal-2026-videomind,
    title = "{V}ideo{M}ind: Thinking in Steps for Long Video Understanding",
    author = "Bhatnagar, Shubhang  and
      Wang, Renxiong  and
      Krishnakumar, Kapil  and
      Ahmadyan, Adel  and
      Lin, Zhaojiang  and
      Mathias, Lambert  and
      Dong, Xin Luna  and
      Damavandi, Babak  and
      Ahuja, Narendra  and
      Moon, Seungwhan",
    editor = {Matusevych, Yevgen  and
      Eryi{\u{g}}it, G{\"u}l{\c{s}}en  and
      Aletras, Nikolaos},
    booktitle = "Proceedings of the 19th Conference of the {E}uropean Chapter of the {A}ssociation for {C}omputational {L}inguistics (Volume 5: Industry Track)",
    month = mar,
    year = "2026",
    address = "Rabat, Morocco",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-eacl/2026.eacl-industry.30/",
    pages = "406--416",
    ISBN = "979-8-89176-384-5",
    abstract = "Multimodal Large Language Models (MLLMs) struggle with Long Video Understanding (LVU) due to their limited context window and the distributed nature of salient information across many redundant frames. To address this, we present VideoMind, a novel training free framework for LVU designed to mimic a human reasoning process. The framework is orchestrated by an MLLM that breaks down a user{'}s query into a series of simpler, actionable sub-queries. For each sub query, the MLLM reconfigures itself by invoking specialized `modes' that are instantiations of the same MLLM, but with appropriately tailored context for the given sub query to extract targeted evidence. After gathering this evidence, the model resumes its role as the orchestrator which evaluates the results and decides if an answer is complete or if it must refine its strategy by engaging further modes with new context. Our specialized operational modes include: 1) a Multi-Scale Temporal Search mode to identify and summarize relevant video sub-snippets at varying time scales, and 2) a Single-Frame Visual Detail mode for precise spatial localization of objects. This dynamic allocation of computation yields state-of-the-art results on the Video-MME, LongVideo, and MLVU benchmarks, achieving 77.6{\%} performance on Video MME using Qwen 2.5 72B (4.8{\%} enhancement) while also yielding a 5{\%} improvement on Llama 4 Scout."
}