Black-Box Face Recovery from Identity Features
This work addresses privacy and security concerns in face recognition systems by enabling black-box recovery, though it is incremental as it builds on existing attack methods.
The authors tackled the problem of reconstructing a face image from only the identity feature vector of a black-box face recognition system, achieving a method that requires significantly fewer queries than the state-of-the-art solution and allows identification by an independent critic network.
In this work, we present a novel algorithm based on an it-erative sampling of random Gaussian blobs for black-box face recovery, given only an output feature vector of deep face recognition systems. We attack the state-of-the-art face recognition system (ArcFace) to test our algorithm. Another network with different architecture (FaceNet) is used as an independent critic showing that the target person can be identified with the reconstructed image even with no access to the attacked model. Furthermore, our algorithm requires a significantly less number of queries compared to the state-of-the-art solution.