Face Recognition System Using Encrypted Feature Vectors for Privacy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional face recognition systems face challenges in balancing user privacy and security, as they often outsource sensitive face data, leading to privacy violations and high maintenance costs, and local solutions are limited by scalability, flexibility, and power consumption.
Innovation Solution
A face recognition system that captures and de-identifies face images using a machine learning model with privacy protection technologies like differential privacy and homomorphic encryption, transforming data into de-identified features for secure identity verification without uploading original images to the cloud, utilizing cloud and edge computing for efficient processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If face data is outsourced to a central server, then identity verification capability is improved, but user privacy protection deteriorates
Solution Approach 1:
The patent extracts only the essential verification features from face images and stores them locally in an encrypted form, separating the verification capability from the original face data. This allows identity verification to be performed without storing or transmitting actual face images to external servers, thus maintaining verification reliability while protecting user privacy.
Solution Approach 2:
The patent introduces encrypted feature vectors as an intermediary between the face image and the verification system. These encrypted features act as a mediator that enables verification without exposing the original face data, allowing the system to verify identity while maintaining privacy protection through cryptographic transformations.
2Object-affected harmful factors
If face data is stored locally, then user privacy is protected to a certain extent, but scalability and flexibility deteriorate
Solution Approach 1:
The patent segments the face recognition system into modular components: local encrypted feature storage, cloud-based verification services, and encrypted communication protocols. This segmentation allows the system to maintain privacy protection locally while enabling scalable cloud-based verification capabilities, improving both privacy protection and system scalability simultaneously.
Solution Approach 2:
The patent creates a universal encrypted feature format that can be stored locally and verified through cloud services, making the system adaptable to multiple deployment scenarios. The encrypted feature vectors serve multiple functions: local storage, cloud verification, and cross-device compatibility, thereby improving scalability without compromising privacy.
3Speed
If conventional face recognition is implemented, then verification speed is improved, but data security and privacy protection deteriorate
Solution Approach 1:
The patent replaces the conventional mechanical approach of storing and transmitting raw face images with a cryptographic system based on encrypted feature vectors. This substitution maintains verification speed through efficient cryptographic operations while eliminating data leakage risks associated with storing and transmitting actual face images.
Data Source
AI summary
A face recognition system and a face recognition method are provided. The face recognition system includes an image capturing device and a processing device. The image capturing device is configured to capture a face image of a user to be recognized, de-identify the face image to obtain de-identified image data, and transform the de-identified image data into multiple de-identified features and output. The processing device is configured to verify an identity of the user to which the de-identified features belong by using a trained machine learning model. The machine learning model is trained by using de-identified features and identities of multiple users registered in advance.