The impact of privacy protection filters on gender recognition

Deep learning-based algorithms have become increasingly ecient in recognition and detection tasks, especially when they are trained on large-scale datasets. Such recent success has led to a speculation that deep learning methods are comparable to or even outperform human visual system in its ability to detect and recognize objects and their features. In this paper, we focus on the specic task of gender recognition in images when they have been processed by privacy protection lters (e.g., blurring, masking, and pixelization) applied at dierent strengths. Assuming a privacy protection scenario, we compare the performance of state of the art deep learning algorithms with a subjective evaluation obtained via crowdsourcing to understand how privacy protection lters aect both machine and human vision.