Face Database Search for Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current solutions for identity fraud detection are inefficient as they fail to capture and extract facial biometrics from ID cards, lack enrollment of ID card faces into databases, and do not utilize metadata for suspicious re-entry checks, making it difficult to distinguish between authentic and counterfeit identification media.
Innovation Solution
A method and system that capture and process selfie and ID card images in real-time, extracting face feature vectors and metadata, searching for nearest neighbors in pre-stored databases, and matching metadata to detect fraudulent actions by generating fraudulent or non-fraudulent indications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing fraud detection solutions are used, then basic identity verification can be performed, but the system fails to capture and extract facial biometrics from ID cards, resulting in poor detection accuracy
Solution Approach 1:
The system segments the fraud detection process into distinct modules: capturing unit for acquiring images, processing unit for extracting facial biometrics and metadata, searching unit for database queries, and decision-making unit for fraud determination. Each module handles specific tasks independently, improving overall detection accuracy while maintaining manageable system complexity through functional decomposition.
Solution Approach 2:
The patent introduces an intermediary database layer that stores extracted facial biometrics and metadata from ID cards. This intermediary structure enables efficient retrieval and comparison of facial features without requiring direct complex analysis of all input data, thereby improving detection accuracy while simplifying the overall processing architecture.
2Reliability
If manual human verification is used, then basic identity checking can be performed, but it becomes difficult and near impossible to stop fraud cases where fraudsters create multiple fake identities
Solution Approach 1:
The system enables self-service fraud detection by automatically extracting facial biometrics from ID card images and comparing them against stored data in the database. This automated process eliminates the need for manual verification while maintaining high reliability in detecting multiple fake identities through automated facial feature matching and metadata analysis.
Solution Approach 2:
The patent replaces manual human verification with an automated electronic system that uses image processing, facial recognition algorithms, and database querying. This substitution of mechanical/manual processes with automated computational methods significantly improves both reliability in detecting fraud patterns and productivity in verification speed.
3Adaptability or versatility
If ID card faces are not enrolled in database, then system remains simple, but search capability for fraudulent actions is limited
Solution Approach 1:
The system performs preliminary action by pre-enrolling and storing facial biometrics and metadata from legitimate ID cards into the database before fraud detection is needed. This advance preparation enables efficient search and comparison operations during actual fraud detection, significantly enhancing search capability while managing database size through selective enrollment of verified identities.
Data Source
AI summary
A system and method for detection of a fraudulent action using face database search and retrieval is given. The method identifies fraudulent action using a combination of images captured in real-time, feature vector extraction, and metadata matching. The method involves capturing, a query selfie image and a query ID card image. Thereafter, face feature vectors from the query images are extracted, and candidate face feature vectors are identified by searching a set of nearest neighbors of the extracted face feature vectors. The method then retrieves pre-stored application IDs associated with the candidate face feature vectors and a pre-stored metadata associated with the candidate face feature vectors. The method then matches the query metadata with the pre-stored metadata and generates either a fraudulent or non-fraudulent indication based on the matching. Thereafter it detects a fraudulent activity based on the fraudulent indication.