Sebastian Raschka • 8/2/2018

Generating Gender-Neutral Face Images with Semi-Adversarial Neural Networks to Enhance Privacy

This article details research into semi-adversarial networks that perturb gender information in face images to prevent automated gender classification. The goal is to enhance user privacy for applications like surveillance or databases, ensuring compliance with regulations like GDPR, while still preserving the image's utility for biometric face recognition tasks. The work is presented as a concise summary of published papers.

0 comments

#privacy #Neural Networks #Face Recognition