Face Identification Grouping and Correction Method
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Face identification algorithms face accuracy issues due to low diversity of images in databases, where each user category often has only a single picture or few pictures, leading to poor retrieval results.
Innovation Solution
A method and device for face identification that involves receiving multiple images, extracting face images, determining initial identification results by matching them to a target image in an image identification library, grouping face images based on features, and correcting identification results to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If face identification is performed using a database with limited images per user category, then the system can operate with simpler data requirements, but the identification accuracy deteriorates due to low diversity of retrievals
Solution Approach 1:
The patent segments the face identification process into multiple stages: initial identification using a standard database, grouping similar faces together, and then performing corrected identification within each group. This segmentation allows the system to handle limited database images per user by breaking down the complex task into manageable steps that improve overall accuracy.
Solution Approach 2:
The patent performs preliminary actions by first extracting face images, then performing initial identification before grouping and corrected identification. This preliminary processing of face images and their initial classification enables the system to prepare data in advance, improving the final identification accuracy even when the original database has limited diversity.
2Measurement precision
If multiple face images are extracted and grouped for corrected identification, then identification accuracy improves, but the processing complexity increases
Solution Approach 1:
The patent merges similar face images into groups based on their initial identification results and feature similarity. By combining multiple face images that belong to the same user or category into unified groups, the system can perform corrected identification that leverages information from multiple sources, improving accuracy without requiring completely separate processing paths for each image.
Solution Approach 2:
The patent introduces an intermediary grouping mechanism that acts as a mediator between the initial identification results and the final corrected identification. This intermediary step of grouping similar faces together simplifies the overall processing by creating organized categories that can be handled systematically, reducing the complexity of processing multiple images individually.
Data Source
AI summary
Aspects of the disclosure can provide method for identifying a face where multiple images to be identified are received. Each of the multiple images includes a face image part. Each face image of face images in the multiple images to be identified is extracted. An initial figure identification result of identifying a figure in the each face image is determined by matching a face in the each face image respectively to a face in a target image in an image identification library. The face images are grouped. A target figure identification result for each face image in each group is determined according to the initial figure identification result for the each face image in the each group.