Development of Privacy Protection Monitoring Systems Using Skeleton Models and Their Evaluation on the Viewpoint of FUBEN-EKI

In this paper, we propose a monitoring system using the pose estimation, which takes account of the protection of privacy. In order to cover a wider area, Region Correspondence Mechanism which associates skeletons of multiple cameras is introduced. In this paper, we examined the accuracy of association of the Region Correspondence Mechanism and verified the confidentiality of the skeleton model. Furthermore, qualitative consideration was made on the benefit of the inconvenience of the proposed system, i.e., FUBEN-EKI. We concluded that there are FUBEN-EKI such as “Enhance awareness”, “Understand systems”, and “Feel at ease”.

[1]  Peter V. Gehler,et al.  DeepCut: Joint Subset Partition and Labeling for Multi Person Pose Estimation , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[3]  Yaser Sheikh,et al.  OpenPose: Realtime Multi-Person 2D Pose Estimation Using Part Affinity Fields , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.