伴随着云机器人的发展，机器人能够参与的场景越来越多，互联网的飞速发展，随之而来的是多用户在不同场景下使用机器人构成的隐私威胁。机器人在不同场景下应拥有不同的权限，在保护用户隐私的同时还可以效率最大化地完成任务。为此，场景理解是解决这一问题的基础，通过预先设定不同场景下各用户所拥有的权利与限制，服务器通过机器人拍摄的图片判断机器人所处的场景，从而赋予各用户不同的权限，使机器人听从用户的指令保护隐私。此举的应用为多用户多机器人系统的安全防护方式提供了一种新的思路。
With the development of cloud robots, robots can participate in more and more different scenes. The rapid development of the Internet is followed by the privacy security issues when multiple users use robots in different scenarios. The robot should have different permissions in different scenarios which can maximize the efficiency and protect the user’s privacy. To solve this problem, scene recognition is one way to solve this problem. By pre-setting the rights and restrictions owned by each user in different scenarios, the server determines the scene in which the robot is located by the picture taken by the robot, thereby giving each user different permissions, which allows the robot to listen to the user’s instructions to protect user’s privacy. The application of this method provides a new way of security protection for the multi-user multi-robot system.
Lin, T.-Y., et al. (2014) Microsoft Coco: Common Objects in Context. European Conference on Computer Vision, Springer, Cham, 740-755. https://doi.org/10.1007/978-3-319-10602-1_48
[9]
Zhou, B.L., et al. (2017) Scene Parsing through ade20k Dataset. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, 21-26 July 2017, 5122-5230. https://doi.org/10.1109/CVPR.2017.544
[10]
Mottaghi, Roozbeh, et al. (2014) The Role of Context for Object Detection and Semantic Segmentation in the Wild. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Columbus, 23-28 June 2014, 891-898. https://doi.org/10.1109/CVPR.2014.119
