叙事对齐的长视频问答
Narrative Aligned Long Form Video Question Answering
AI总结：
本文提出NA-VQA基准，评估长视频中的深度时间与叙事推理能力，通过88部完整电影和4.4K开放性问题测试模型整合分散叙事信息的能力，提出Video-NaRA框架提升远距离推理性能。
摘要：
近期多模态大语言模型（MLLMs）的进步推动了长视频推理基准的兴起。然而，现有基准多依赖局部线索，无法捕捉叙事推理能力，即跟踪意图、连接远距离事件并重建因果链。我们引入NA-VQA基准，包含88部完整电影和4.4K开放性问题-答案对，每个问题基于多个证据片段标注为短、中或远距离，以评估长距离依赖。通过要求生成多场景答案，NA-VQA测试模型是否能整合分散的叙事信息而非依赖浅层模式匹配。为解决现有方法的局限性，我们提出Video-NaRA框架，构建事件级链并存储在结构化记忆中供推理检索。实验表明，最先进的MLLMs在需要远距离证据的问题上表现不佳，凸显了显式叙事建模的必要性。Video-NaRA将长距离推理性能提升高达3%，证明其在处理复杂叙事结构的有效性。
英文摘要：
Recent progress in multimodal large language models (MLLMs) has led to a surge of benchmarks for long-video reasoning. However, most existing benchmarks rely on localized cues and fail to capture narrative reasoning, the ability to track intentions, connect distant events, and reconstruct causal chains across an entire movie. We introduce NA-VQA, a benchmark designed to evaluate deep temporal and narrative reasoning in long-form videos. NA-VQA contains 88 full-length movies and 4.4K open-ended question-answer pairs, each grounded in multiple evidence spans labeled as Short, Medium, or Far to assess long-range dependencies. By requiring generative, multi-scene answers, NA-VQA tests whether models can integrate dispersed narrative information rather than rely on shallow pattern matching. To address the limitations of existing approaches, we propose Video-NaRA, a narrative-centric framework that builds event-level chains and stores them in a structured memory for retrieval during reasoning. Extensive experiments show that state-of-the-art MLLMs perform poorly on questions requiring far-range evidence, highlighting the need for explicit narrative modeling. Video-NaRA improves long-range reasoning performance by up to 3 percent, demonstrating its effectiveness in handling complex narrative structures. We will release NA-VQA upon publication.
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。
