Generation of Gradient-Preserving Images allowing HOG Feature Extraction

2021/04/03 by Masaki Kitayama, Kitayama, Masaki, Hitoshi Kiya +1
Computer Science · #Face recognition and analysis #Biometric Identification and Security #Advanced Steganography and Watermarking Techniques

paper · pdf · doi:10.48550/arxiv.2104.01350

Abstract

In this paper, we propose a method for generating visually protected images, referred to as gradient-preserving images. The protected images allow us to directly extract Histogram-of-Oriented-Gradients (HOG) features for privacy-preserving machine learning. In an experiment, HOG features extracted from gradient-preserving images are applied to a face recognition algorithm to demonstrate the effectiveness of the proposed method.

Related

• HOG feature extraction from encrypted images for privacy-preserving machine learning
• Privacy-Preserving Face Recognition Using Trainable Feature Subtraction
• MediaEval 2019: Concealed FGSM Perturbations for Privacy Preservation
• Inverting Gradients -- How easy is it to break privacy in federated learning?
• Enhancing Generalization of Invisible Facial Privacy Cloak via Gradient Accumulation