VHAKG: A Multi-modal Knowledge Graph Based on Synchronized Multi-view Videos of Daily Activities
This work addresses the need for structured video data resources for researchers in AI and computer vision, though it is incremental as it builds on existing MMKG concepts with a specific focus on daily activities.
The paper tackles the problem of constructing multi-modal knowledge graphs (MMKGs) for videos of daily activities, which are lacking, by building an MMKG based on synchronized multi-view simulated videos that includes event-centric knowledge and fine-grained frame-by-frame details, and demonstrates its utility in benchmarking vision-language models.
Multi-modal knowledge graphs (MMKGs), which ground various non-symbolic data (e.g., images and videos) into symbols, have attracted attention as resources enabling knowledge processing and machine learning across modalities. However, the construction of MMKGs for videos consisting of multiple events, such as daily activities, is still in the early stages. In this paper, we construct an MMKG based on synchronized multi-view simulated videos of daily activities. Besides representing the content of daily life videos as event-centric knowledge, our MMKG also includes frame-by-frame fine-grained changes, such as bounding boxes within video frames. In addition, we provide support tools for querying our MMKG. As an application example, we demonstrate that our MMKG facilitates benchmarking vision-language models by providing the necessary vision-language datasets for a tailored task.