Generation of Gradient-Preserving Images allowing HOG Feature Extraction
This addresses privacy concerns in machine learning for applications like face recognition, but appears incremental as it builds on existing HOG feature methods.
The paper tackles the problem of generating visually protected images that preserve gradients, enabling direct extraction of HOG features for privacy-preserving machine learning, and demonstrates effectiveness by applying these features to a face recognition algorithm.
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.