arXiv 2510.177222026-01-08cs.CVcs.AI

MT-Video-Bench: 一个多轮对话中评估多模态大语言模型的综合视频理解基准

MT-Video-Bench: A Holistic Video Understanding Benchmark for Evaluating Multimodal LLMs in Multi-Turn Dialogues

Fudan University（复旦大学）；NJU-LINK Team, Nanjing University（南京大学NJU-LINK团队）；University of Science and Technology of China（中国科学技术大学）；Kuaishou Technology（快手科技）

Yaning Pan, Qianqian Xie, Guohui Zhang, Zekun Wang, Yongqian Wen, Yuanxing Zhang, Haoxuan Hu, Zhiyu Pan, Yibing Huang, Zhidong Gan, Yonghong Lin, An Ping, Shiha… 展开作者

Yaning Pan, Qianqian Xie, Guohui Zhang, Zekun Wang, Yongqian Wen, Yuanxing Zhang, Haoxuan Hu, Zhiyu Pan, Yibing Huang, Zhidong Gan, Yonghong Lin, An Ping, Shihao Li, Yanghai Wang, Tianhao Peng, Jiaheng Liu

备注：: Project Website: https://github.com/NJU-LINK/MT-Video-Bench

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

MT-Video-Bench通过评估多轮对话中多模态大语言模型的6项核心能力，揭示其在处理多轮视频对话任务中的性能差异和局限性。

摘要：

最近多模态大语言模型（MLLMs）的发展显著提升了AI对视觉模态的理解能力。然而，现有评估基准仍然局限于单轮问答，忽略了现实场景中多轮对话的复杂性。为弥合这一差距，我们引入MT-Video-Bench，这是一个用于评估多轮对话中MLLMs的综合视频理解基准。具体而言，我们的MT-Video-Bench主要评估6项核心能力，专注于感知力和交互性，涵盖来自不同领域的1000个精心挑选的多轮对话。这些能力严格与现实应用对齐，如互动体育分析和多轮视频智能教学。通过MT-Video-Bench，我们广泛评估了各种最先进的开源和闭源MLLMs，揭示了它们在处理多轮视频对话中的显著性能差异和局限性。该基准将公开发布以促进未来研究。

英文摘要：

The recent development of Multimodal Large Language Models (MLLMs) has significantly advanced AI's ability to understand visual modalities. However, existing evaluation benchmarks remain limited to single-turn question answering, overlooking the complexity of multi-turn dialogues in real-world scenarios. To bridge this gap, we introduce MT-Video-Bench, a holistic video understanding benchmark for evaluating MLLMs in multi-turn dialogues. Specifically, our MT-Video-Bench mainly assesses 6 core competencies that focus on perceptivity and interactivity, encompassing 1,000 meticulously curated multi-turn dialogues from diverse domains. These capabilities are rigorously aligned with real-world applications, such as interactive sports analysis and multi-turn video-based intelligent tutoring. With MT-Video-Bench, we extensively evaluate various state-of-the-art open-source and closed-source MLLMs, revealing their significant performance discrepancies and limitations in handling multi-turn video dialogues. The benchmark will be publicly available to foster future research.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。