DETECTRON2 Object Detection Dataset

该数据集专为在受控环境中进行目标检测而设计，尤其关注家庭场景，如厨房、花园和食品储藏室。它采用了Detectron2目标检测框架的高级扩展，包含了超过21,000个类别。此外，该数据集支持对Robo-CSK-Organizer和ChatGPT基准系统进行评估，使得可以在基于上下文识别的目标放置方面进行比较。规模上，该数据集涵盖了21,000多个类别，任务是对检测到的上下文进行目标分类和组织。

This dataset is specifically designed for object detection in controlled environments, with a particular focus on household scenarios such as kitchens, gardens, and pantries. It adopts the advanced extension of the Detectron2 object detection framework and contains over 21,000 categories. Additionally, this dataset supports evaluation on the Robo-CSK-Organizer and ChatGPT benchmark systems, enabling comparative analyses of object placement based on context recognition. In terms of scale, this dataset encompasses more than 21,000 categories, and its core task is to perform object classification and organization on detected contexts.