Real-World Robotics Lab Dataset

该数据集包含了在杂乱无章的机器人实验室环境内收集的10,132个图像帧，这些图像覆盖了22小时的时间跨度。数据集中不仅包含了静态物体，还有动态物体，例如各种杯子和大瓶子的位置变化。这个数据集在一个光照恒定的非受控环境中收集，并用于评估D3A算法在物体检索方面的表现。此外，该数据集的规模达到了10,132个图像帧，其任务重点在于物体检索以及时空表征。

This dataset includes 10,132 image frames collected in a cluttered robotic laboratory environment, covering a 22-hour time span. It contains both static objects and dynamic objects, such as the positional variations of various cups and large bottles. This dataset was gathered in an uncontrolled environment with constant lighting, and is employed to evaluate the performance of the D3A algorithm in object retrieval. Furthermore, with a total of 10,132 image frames, the key tasks of this dataset focus on object retrieval and spatiotemporal representation.