Saved in:
Bibliographic Details
Main Authors: Egami, Shusaku, Ugai, Takahiro, Htun, Swe Nwe Nwe, Fukuda, Ken
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2408.14895
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916372838088704
author Egami, Shusaku
Ugai, Takahiro
Htun, Swe Nwe Nwe
Fukuda, Ken
author_facet Egami, Shusaku
Ugai, Takahiro
Htun, Swe Nwe Nwe
Fukuda, Ken
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2408_14895
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle VHAKG: A Multi-modal Knowledge Graph Based on Synchronized Multi-view Videos of Daily Activities
Egami, Shusaku
Ugai, Takahiro
Htun, Swe Nwe Nwe
Fukuda, Ken
Artificial Intelligence
Computation and Language
Computer Vision and Pattern Recognition
68T30
I.2.4; H.5.1
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.
title VHAKG: A Multi-modal Knowledge Graph Based on Synchronized Multi-view Videos of Daily Activities
topic Artificial Intelligence
Computation and Language
Computer Vision and Pattern Recognition
68T30
I.2.4; H.5.1
url https://arxiv.org/abs/2408.14895