Image Verification Device Using Relative Evaluation Scores
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional face authentication systems face challenges in accurately distinguishing between individuals with similar facial features, leading to erroneous authentication, particularly when threshold values are difficult to set for relatives like brothers or children.
Innovation Solution
An image verification device that calculates an integrated score by combining a verification score representing the degree of approximation between an input image and registered images with a relative evaluation score, which assesses the similarity between registered images, to improve authentication accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If threshold values are set for authentication evaluation, then authentication accuracy can be improved, but it becomes difficult to set appropriate threshold values for persons with similar faces
Solution Approach 1:
The patent changes the evaluation parameter from absolute threshold-based scoring to relative differential scoring. Instead of comparing authentication scores against fixed threshold values, the system calculates the difference between the highest and second highest authentication scores to determine authentication results. This parameter transformation eliminates the need for difficult threshold setting while maintaining authentication accuracy.
2Productivity
If simple comparison with registered images is used, then authentication process is fast and simple, but erroneous authentication occurs between persons with similar faces
Solution Approach 1:
The patent applies partial action by calculating authentication scores only against the top N registered images (e.g., top 5 or top 10) rather than all registered images. This partial comparison maintains fast processing speed while improving accuracy by focusing on the most likely matches. The system then uses differential scoring among these top candidates to prevent erroneous authentication.
Solution Approach 2:
The patent introduces dynamic evaluation by calculating differential scores based on the relative differences between authentication scores rather than using static threshold values. The authentication decision dynamically adapts to the specific case by considering the score gap between candidates, making the system more reliable for persons with similar faces while maintaining efficiency.
3Reliability
If grouping of similar face registered data is performed, then threshold setting becomes possible, but appropriate threshold values are still difficult to set and require trial and error
Solution Approach 1:
The patent enables the system to self-adjust and self-evaluate by automatically calculating differential scores between authentication results. Instead of requiring manual threshold setting or grouping operations, the system autonomously determines authentication outcomes by comparing score differences, eliminating the need for operator intervention in threshold configuration while maintaining high reliability.
Data Source
AI summary
An image verification device that checks an input image obtained by photographing an object to be checked against a registered image database, wherein, in the registered image database, an amount of feature of an image obtained by photographing an object is registered as a registered image, and the registered image includes registered images registered with respect to a plurality of objects, has a verification score calculating unit that calculates a verification score serving as a score representing a degree of approximation between the objects indicated by the registered images and the object of the input image by using the amount of feature of the input image and the amounts of feature of the registered images, and a relative evaluation score calculating unit.