underctrl/toy_pickup_and_place

这是一个关于机器人操作的数据集，包含50个剧集，共计29850帧，1个任务，100个视频。数据集分为训练集，其比例为0到50。每个视频包含两种类型的图像：右侧摄像头图像和手机图像，均为480x640的3通道视频流，帧率为30fps，采用h264编码。数据集还包括机器人的动作和状态信息，数据类型为float32，共有7个维度，分别对应不同的机器人关节角度。所有数据以Parquet格式存储，并提供了相应的视频文件。

This is a dataset about robotic operations, containing 50 episodes, a total of 29,850 frames, 1 task, and 100 videos. The dataset is split into a training set with a ratio of 0 to 50. Each video includes two types of images: right camera images and phone images, both of which are 480x640 3-channel video streams with a frame rate of 30fps encoded in h264. The dataset also includes robotic action and state information, with data types of float32 and 7 dimensions, corresponding to different robotic joint angles. All data is stored in Parquet format, and associated video files are provided.