构建了基于第一视角动作场景图的视频问答基准，用于评估模型对时空关系的理解能力。
EASG-Bench: Video Q&A Benchmark with Egocentric Action Scene Graphs
- 基于动态场景图生成带时空定位的问答对，刻画角色、动作与物体间复杂关系。
- 发现语言模型和视频大模型在时序顺序类问题上表现差距明显，暴露长上下文理解短板。
- 适合关注第一视角视频理解、多模态推理与评测基准研究的学者使用。
我们提出了EASG-Bench，一个针对第一视角视频的问答基准，其问答对源自捕捉演员、动作与物体之间复杂关系的时空定位动态场景图。我们设计了一套系统化的评估框架，并在此基准上评测了几种仅语言模型与视频大语言模型（video-LLMs）。实验发现，语言模型与视频大模型在涉及时序顺序的问题上存在显著性能差距，揭示了长上下文视频理解领域的研究空白。为促进结果可复现并推动后续研究，本基准及配套代码已开源至GitHub：https://github.com/fpv-iplab/EASG-bench。
原文摘要 · Abstract (English)
We introduce EASG-Bench, a question-answering benchmark for egocentric videos where the question-answering pairs are created from spatio-temporally grounded dynamic scene graphs capturing intricate relationships among actors, actions, and objects. We propose a systematic evaluation framework and evaluate several language-only and video large language models (video-LLMs) on this benchmark. We observe a performance gap in language-only and video-LLMs, especially on questions focusing on temporal ordering, thus identifying a research gap in the area of long-context video understanding. To promote the reproducibility of our findings and facilitate further research, the benchmark and accompanying code are available at the following GitHub page: https://github.com/fpv-iplab/EASG-bench.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。