N-UCLA

该数据集名为N-UCLA，包含了1494个序列，涵盖了10个动作类别，这些动作由10位受试者执行。值得注意的是，每位受试者对每个动作只执行一次。整个数据集的规模为1494个序列，其任务是进行动作识别。Named N-UCLA, this dataset comprises 1494 sequences covering 10 action categories, which are performed by 10 subjects. Notably, each subject performs each action only once. The total scale of this dataset is 1494 sequences, and its core task is action recognition.

搜集汇总

背景与挑战

背景概述

N-UCLA数据集是一个多视角3D动作数据集，包含RGB、深度和人体骨骼数据，涵盖10个动作类别，每个动作由10个演员执行，数据从不同视角捕获。