Vehicle Face Authentication Correction via Failure Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional face authentication systems face challenges in accurately authenticating users when the face image obtained during authentication differs from the pre-registered authentication face image due to factors like vehicle attributes and environmental conditions, leading to incorrect authentication.
Innovation Solution
An information processing apparatus and method that acquires user identification information, compares it with a pre-registered authentication face image and a real-time face image captured from outside the vehicle, and generates first information for subsequent authentications when the initial authentication fails, adjusting the authentication process to account for discrepancies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If face authentication is performed by comparing a pre-registered authentication face image with a real-time face image captured from outside the vehicle, then authentication speed is improved, but authentication accuracy deteriorates due to differences in image conditions caused by vehicle attributes and environmental factors
Solution Approach 1:
The system performs preliminary actions by capturing multiple face images at different timings before authentication, and pre-processes these images to generate correction information. This preparation work is done in advance so that during actual authentication, the system can quickly apply pre-computed corrections without sacrificing accuracy.
Solution Approach 2:
The system introduces feedback by comparing the captured face image with the authentication face image, detecting differences caused by vehicle attributes and environmental conditions, and then correcting the captured image based on this feedback before final authentication. This closed-loop approach maintains accuracy while preserving fast authentication.
2Measurement precision
If the authentication system requires multiple face images captured at different timings, then authentication accuracy is improved by selecting the best match, but processing time and system complexity increase
Solution Approach 1:
The system implements self-service by automatically capturing multiple face images, comparing them, selecting the appropriate one, and generating correction information without requiring external intervention or complex manual processing. The system serves itself by autonomously managing the entire multi-image authentication workflow.
Solution Approach 2:
The system changes parameters by capturing face images at different timings and under different conditions, then selects the optimal image based on comparison results. By varying temporal and environmental parameters, the system improves authentication accuracy while using automated selection to avoid excessive complexity.
3Ease of operation
If face images are captured through vehicle windows, then convenience is improved as users can be authenticated while remaining in the vehicle, but image quality deteriorates due to reflections, refractions, and color distortions
Solution Approach 1:
The system converts the harmful effects of window glass (reflections, refractions, color distortions) into beneficial correction information. By deliberately capturing images through the window and analyzing the distortions, the system generates correction data that compensates for these effects, ultimately improving authentication accuracy while maintaining user convenience.
Solution Approach 2:
The system uses the vehicle window as an intermediary that provides useful information about environmental conditions and image distortions. By analyzing how the window affects the captured image, the system generates correction information that compensates for the window's detrimental effects, turning the intermediary from a source of error into a source of correction data.
Data Source
AI summary
A controller is provided which acquires user identification information that is identification information of a user of a vehicle; acquires an authentication face image that is an image associated with the user identification information and is an image of a face for authentication of the user of the vehicle; acquires a user's face image that is an image generated by photographing the user of the vehicle with a camera disposed outside the vehicle; performs face authentication based on a comparison between the authentication face image and the user's face image; and acquires, from the user's face image, first information to be used in the next and subsequent face authentications in response to the face authentication not being successful.