Ambiguous Person Detection via Feature Degrading
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current person recognition technologies, particularly those using neural networks, raise concerns about the far-reaching use of personal data and violate data protection laws by enabling unique identification of individuals without consent, necessitating a method that ensures ambiguous identification to respect privacy.
Innovation Solution
A method that classifies persons in monitored areas by extracting and degrading image features from multiple cameras, ensuring ambiguity in classification to prevent unique identification, using neural networks to output multiple comparison persons and forming collective trajectories while maintaining data protection compliance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If person recognition technology uses neural networks to detect and track persons, then detection accuracy and tracking reliability are improved, but data protection laws are violated by enabling unique identification of individuals
Solution Approach 1:
The patent extracts only the necessary classification features from person images while deliberately excluding biometric data that would enable unique identification. The system extracts values for classification features (such as clothing color, body shape, size) without processing facial recognition or other identifying biometric information, thus achieving person detection while protecting privacy.
Solution Approach 2:
The patent changes the parameters used for person classification by deliberately degrading image quality and limiting the types of features extracted. Instead of using high-resolution images with biometric data, the system uses degraded images with non-identifying classification features, thereby maintaining detection capability while preventing unique identification.
2Object-affected harmful factors
If the classification is made ambiguous to prevent unique identification, then data protection compliance is improved, but the ability to uniquely identify the search person deteriorates
Solution Approach 1:
The patent applies partial action by outputting multiple comparison persons (at least two) instead of a single identification result. This provides sufficient information for search purposes while deliberately avoiding excessive precision that would enable unique identification. The system extracts classification features from multiple persons and outputs them as comparison results, ensuring privacy protection while maintaining search utility.
Solution Approach 2:
The patent creates copies of classification data by generating multiple comparison person profiles with similar classification features. Instead of providing one precise identification, the system provides multiple comparable profiles that contain sufficient information for search purposes without enabling unique identification, thus balancing privacy protection with search effectiveness.
3Adaptability or versatility
If multiple comparison persons are output from the comparison, then the search space is expanded and comprehensive results are generated, but the precision of identifying the specific search person is reduced
Solution Approach 1:
The patent deliberately outputs more comparison persons than the minimum single identification would provide, expanding the search coverage to include multiple potential matches. This partial action approach provides sufficient information for comprehensive search results while deliberately stopping short of providing precise unique identification, thus balancing search versatility with privacy protection.
Data Source
AI summary
A method for detecting comparison persons 7 to a search person 4, wherein a plurality of classification persons 3 is classified by extracting values W1,W2,W3 for classification features K1,K2,K3 from classification images 2 of the classification persons 3, the classification being ambiguous in such a way that the classification does not enable a unique identification of any of the classification persons 3, wherein during a search for a search person 4 using a search image 5 by a comparison of values of search features from the search image 5 with values W1,W2,W3 of classification features K1,K2,K3, at least two classification persons 3 are output as comparison persons 7.