Face Image Quality Assessment Using Pose and Size Metrics

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current face recognition technologies face challenges in accurately evaluating face image quality, leading to high false recognition rates due to poor image quality, particularly issues with pose angles and face size, which are not adequately addressed by existing evaluation methods.

Innovation Solution

A method and apparatus for determining face image quality by obtaining pose angle and size information, using face detection bounding boxes and key point coordinates to calculate quality scores based on yaw and pitch angles, and face size, thereby objectively evaluating face image quality and improving recognition accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If face recognition is performed on images with varying pose angles and sizes, then the system can handle diverse real-world scenarios, but the recognition accuracy deteriorates due to poor image quality

Engineering Contradiction:

Improvehandling diverse scenesVSAvoidrecognition accuracy

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system performs quality assessment before face recognition to pre-filter unsuitable images. By evaluating pose angles and face sizes in advance, the system identifies and excludes low-quality images that would compromise recognition accuracy, thus maintaining high reliability while handling diverse scenarios.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The quality assessment mechanism provides feedback on image suitability based on pose angles and face sizes. This feedback loop allows the system to adjust its processing by filtering out images that do not meet quality thresholds, thereby maintaining recognition accuracy across varied real-world conditions.

Inventive Principle:

Principle #23Feedback

2Device complexity

If traditional quality assessment methods are used, then the evaluation process is simple, but the accuracy of quality evaluation is insufficient leading to high false recognition rates

Engineering Contradiction:

Improveevaluation process simplicityVSAvoidquality evaluation accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The system introduces specific parameters (pose angles and face sizes) to enhance quality evaluation accuracy. By measuring and analyzing these parameters, the system achieves more precise quality assessment that directly addresses the causes of false recognition, moving beyond simple evaluation methods.

Inventive Principle:

Principle #35Parameter changes

3Reliability

If all captured images are processed for face recognition, then no images are lost, but processing time increases and operational efficiency decreases

Engineering Contradiction:

Improvecompleteness of processingVSAvoidoperational efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system extracts and evaluates key parameters (pose angles and face sizes) to identify and separate high-quality images from low-quality ones. By extracting only the necessary quality indicators, the system filters out unsuitable images before recognition processing, maintaining completeness for high-quality images while improving operational efficiency through selective processing.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS11182589B2Methods and apparatuses for determining face image quality, electronic devices, and computer storage media

Publication Date: 2021.11.23 SHENZHEN SENSETIME TECH CO LTD

AI summary

A methods for determining face image quality includes: obtaining pose angle information and/or size information of a face in an image; and obtaining quality information of the face in the image on the basis of the pose angle information and/or the size information of the face.