CLAD (Complex and Long Activities Dataset)

CLAD（Compled and Long Activities Dataset）是一个活动数据集，展示了复杂、时间扩展的人类活动和行动的真实和多样化场景。该数据集由一组演员以自然且无脚本的方式进行日常活动的视频组成。该数据集使用静态 Kinect 2 传感器记录，该传感器通常用于许多机器人平台。该数据集包括 RGB-D 图像、点云数据、自动生成的骨架轨迹以及众包注释。CLAD (Complex and Long Activities Dataset) is an activity dataset that showcases realistic and diverse scenarios of complex, time-extended human activities and actions. This dataset consists of videos of daily activities performed by a group of actors in a natural, unscripted manner. It was recorded using a static Kinect 2 sensor, which is commonly utilized on many robotic platforms. This dataset includes RGB-D images, point cloud data, automatically generated skeleton trajectories, and crowdsourced annotations.

搜集汇总

背景与挑战

背景概述

CLAD是一个复杂且长时间的人类活动数据集，包含演员自然表演日常活动的视频，使用Kinect 2传感器记录，提供RGB-D图像、点云、骨架轨迹和众包注释。该数据集由利兹大学于2017年发布，适用于视频分类和行为识别等任务。