Sims4Action

该数据集是为了通过学习生活模拟视频游戏来识别家庭中的人类行为而设计的，特别是使用了《模拟人生4》来创建。数据集包含了平衡表示各项活动的视频，这些视频从不同的角度拍摄，并根据主体划分为训练集、验证集和测试集。规模上，该数据集涵盖了942个视频，涉及十个不同行为，且主体多样。其任务是识别日常生活中的活动（Adl）。This dataset is designed for recognizing human household behaviors by learning from life simulation video games, and it is specifically constructed using The Sims 4. The dataset includes balanced videos representing various daily activities, which are filmed from multiple perspectives and split into training, validation and test sets based on the subjects. In terms of scale, this dataset contains 942 videos covering ten distinct activity categories, with diverse participating subjects. The target task of this dataset is to recognize Activities of Daily Living (ADLs).