LAUSR

When taking photos of a suspicious missing child in the street and posting them to the social network is becoming a feasible way to find missing children, the exposure of photos may cause privacy issues. To address this problem, we propose Double-Blinded Finder, an efficient and double-blinded system for finding missing children via low-dimensional multi-attribute representation of child face and blind face matching. To obtain enough knowledge for representing child faces, we build the Labeled Child Face in the Wild dataset, which contains 60K Internet images with 6K unique identities. Based on this dataset, we further train a multi-task deep face model to describe a child face as a 128d fixed-point feature vector and extensive gender and age attributes. Using the keys generated from face descriptors, the face photos from the social network and face representation from the parent(s) of missing children are both encrypted. In addition, we devise inner-production encryption to run blind face matching in the public cloud. In this manner, Double-Blinded Finder can provide efficient face matching while protecting the privacies of both sides: (1) the suspicious missing children side for avoiding the invasion of the human rights, and (2) the true missing children side for preserving the secondary victimization. The experiments show that our system can achieve practical performance of child face matching with negligible leakage of privacy.