In-Vehicle Face Morphing for Privacy-Safe Occupant Services
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle service providing systems face challenges in transmitting occupant data while ensuring personal information protection, particularly face images, due to legal restrictions, which can hinder accurate service provision.
Innovation Solution
The system employs face morphing technology using point group and patch data to generate a morphed face image, transmitting these instead of actual occupant images, along with classification data that does not uniquely identify individuals, allowing for service provision while protecting personal information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If actual occupant face images are transmitted to the server apparatus, then accurate service provision is enabled, but personal information protection is compromised
Solution Approach 1:
The system creates a morphed face image that copies the essential facial characteristics and expressions of the occupant without transmitting the actual face image. This copy retains sufficient information for service provision (emotion recognition, occupancy detection) while eliminating personally identifiable features through morphing with sample face images
Solution Approach 2:
The morphed face image acts as an intermediary between the actual occupant face and the server processing system. It mediates by preserving service-critical information (facial structure, expressions) while filtering out personal identification information, allowing accurate service provision without direct exposure of personal data
2Object-affected harmful factors
If point group data and patch data are generated and transmitted instead of captured images, then personal information privacy is maintained, but data processing complexity increases
Solution Approach 1:
The face image processing is segmented into distinct components: point group data representing facial landmarks and patch data representing facial regions. This segmentation allows the system to transmit only essential structural information rather than complete images, reducing privacy risks while organizing data for efficient morphing processing
Solution Approach 2:
The system transforms the captured face image into different data parameters (point coordinates, patch descriptors) that represent facial geometry and features. This parameter transformation enables the generation of morphed images by manipulating these parameters with sample face data, achieving privacy protection through mathematical transformation rather than direct image transmission
Data Source
AI summary
A vehicle service providing system includes a vehicle and a server apparatus. The vehicle includes an in-vehicle camera and one or more vehicle processors. The server apparatus includes an output unit, one or more server memories, and one or more server processors. The one or more vehicle processors transmit, to the server apparatus, point group data and patch data regarding the face of an occupant, together with classification data regarding the occupant or the vehicle with which the occupant is uniquely unidentifiable. The one or more server processors carry out morphing with the use of point group data and patch data regarding a selected one of sample face images, and the point group data and the patch data generated regarding the occupant, to generate a morphed face image, and allow the output unit to output the morphed face image, instead of a face image of the occupant.