HVU
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/holistic-video-understanding/hvu-dataset
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是一个大规模的基准数据集，旨在解决涉及多个语义方面的多标签和多任务视频理解问题。它覆盖了广泛的语义方面，包括场景、物体和动作。该数据集的任务是进行多任务视频理解。
This dataset is a large-scale benchmark dataset designed to address multi-label and multi-task video understanding problems involving multiple semantic aspects. It covers a wide range of semantic aspects including scenes, objects and actions. The task of this dataset is multi-task video understanding.
提供机构：
University of California, Berkeley
