Overview

In the past few years, robotic technology has developed rapidly, and the practical application of intelligent robots such as the Cloud Shadow X20 has become increasingly widespread. For robotic applications, model training requires a large amount of data, and the ability of models to learn throughout their lifetime has become increasingly important due to the complex and ever-changing real environment. However, currently there are relatively few open-source datasets in the field of robotics, making it difficult for developers in the robotics field to obtain high-quality data for model training.

Therefore, we have jointly released the Cloud-Robotics dataset on KubeEdge-Ianvs to provide better support for robot learning. The Cloud-Robotics dataset not only provides high-quality data support for robotic tasks but also offers more possibilities for testing the lifelong learning ability of robot application models. We believe that this dataset will become one of the important tools in the field of robot learning and application development.

Cloud-Robotics is a multimodal robot dataset collected by an intelligent robotic dog in the Huawei Shenzhen Industrial Park. Currently, the outdoor image semantic segmentation dataset of Cloud-Robotics has been open-sourced. In the future, we will open-source other modalities of Cloud-Robotics datasets, including but not limited to point clouds and IMU, etc.

图片描述

Examples

Some examples in Cloud-Robotics dataset

Annoucement

- The Cloud-Robotics official website has been online! (2023/06/30)
- The Cloud-Robotics Semantic Segmentation dataset has been released! (2023/06/30)

Citation

 
    @online{Cloud-Robotics,
        title={Cloud Robotics},
        author={Hu, Shijing and Mao, Sitong and Luo, Siqi and Huang, Zhiwei and Zheng, Zimu and Pu, Jie and Wang, Feng},
        url={https://kubeedge-ianvs.github.io/}
        year={2023}
    }