Kinetics Human Action Video Dataset

Kinetics数据集是由深智研究院创建的大规模人类动作视频数据集，包含400个人类动作类别，每个类别至少有400个视频片段，总计超过24万个训练视频。数据集的视频主要来源于YouTube，涵盖了从个人动作到人与对象互动的广泛动作类型。创建过程中，数据集旨在捕捉真实世界中的复杂动作和场景，以推动视频动作识别技术的发展。该数据集适用于深度学习和计算机视觉领域，特别是用于解决视频动作识别问题，帮助模型理解和分类视频中的动作。

The Kinetics Dataset is a large-scale human action video dataset created by the Shenzhi Research Institute. It includes 400 human action categories, with each category containing at least 400 video clips, and the total number of training videos exceeds 240,000. The videos in this dataset are mainly sourced from YouTube, covering a wide range of action types from individual movements to human-object interactions. During the development process, this dataset was designed to capture complex real-world actions and scenarios, so as to promote the development of video action recognition technologies. This dataset is applicable to the fields of deep learning and computer vision, and is particularly used to address video action recognition tasks, helping models understand and classify actions in videos.