{"ID":2851228,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.20622","arxiv_id":"2510.20622","title":"SeViCES: Unifying Semantic-Visual Evidence Consensus for Long Video Understanding","abstract":"Long video understanding remains challenging due to its complex, diverse, and temporally scattered content. Although video large language models (Video-LLMs) can process videos lasting tens of minutes, applying them to truly long sequences is computationally prohibitive and often leads to unfocused or inconsistent reasoning. A promising solution is to select only the most informative frames, yet existing approaches typically ignore temporal dependencies or rely on unimodal evidence, limiting their ability to provide complete and query-relevant context. We propose a Semantic-Visual Consensus Evidence Selection (SeViCES) framework for effective and reliable long video understanding. SeViCES is training-free and model-agnostic, and introduces two key components. The Semantic-Visual Consensus Frame Selection (SVCFS) module selects frames through (1) a temporal-aware semantic branch that leverages LLM reasoning over captions, and (2) a cluster-guided visual branch that aligns embeddings with semantic scores via mutual information. The Answer Consensus Refinement (ACR) module further resolves inconsistencies between semantic- and visual-based predictions by fusing evidence and constraining the answer space. Extensive experiments on long video understanding benchmarks show that SeViCES consistently outperforms state-of-the-art methods in both accuracy and robustness, demonstrating the importance of consensus-driven evidence selection for Video-LLMs.","short_abstract":"Long video understanding remains challenging due to its complex, diverse, and temporally scattered content. Although video large language models (Video-LLMs) can process videos lasting tens of minutes, applying them to truly long sequences is computationally prohibitive and often leads to unfocused or inconsistent reas...","url_abs":"https://arxiv.org/abs/2510.20622","url_pdf":"https://arxiv.org/pdf/2510.20622v1","authors":"[\"Yuan Sheng\",\"Yanbin Hao\",\"Chenxu Li\",\"Shuo Wang\",\"Xiangnan He\"]","published":"2025-10-23T14:55:28Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}