Obscured Face Swinging Direction Calculation Using 3D Model Anchors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current face detection technologies fail to accurately calculate the swinging direction of a human face when the face is obscured by objects like masks, sunglasses, or hats, as they either cannot detect the face correctly or cannot perform this calculation.
Innovation Solution
A device and method that combines non-obscured and obscured face detection technologies to identify feature anchor points, adjusts a three-dimensional model, and uses these points to calculate the swinging direction of an obscured human face by selecting candidate feature anchor points and translating them to determine updated coordinates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If non-obscured face detection technology is used, then the swinging direction can be calculated accurately for visible faces, but the detection fails when the face is obscured by masks, sunglasses, or hats
Solution Approach 1:
The patent combines non-obscured face detection technology and obscured face detection technology into a unified system. The processor integrates results from both detection methods to determine feature anchor points, allowing accurate swinging direction calculation whether the face is obscured or not.
Solution Approach 2:
The patent introduces a three-dimensional model as an intermediary to bridge the gap between obscured face detection and swinging direction calculation. Candidate feature anchor points from obscured detection are mapped to corresponding points on the three-dimensional model to enable accurate swinging direction computation even when the face is partially hidden.
2Reliability
If obscured face detection technology is used, then obscured faces can be detected, but the swinging direction cannot be calculated
Solution Approach 1:
The three-dimensional model serves as an intermediary that enables swinging direction calculation from obscured face detection results. By mapping candidate feature anchor points to the three-dimensional model, the system recovers spatial orientation information that would otherwise be lost in obscured conditions.
Solution Approach 2:
The patent transforms the output of obscured face detection (candidate feature anchor points) into swinging direction parameters by mapping them to a three-dimensional model. This parameter transformation enables the calculation of swinging direction from detection data that was not originally designed for this purpose.
3Device complexity
If only one detection technology is used, then the system complexity is low, but the system cannot handle both obscured and non-obscured faces effectively
Solution Approach 1:
The patent merges non-obscured and obscured face detection technologies into a single integrated system. The processor coordinates both detection methods, selecting or combining results based on whether the face is obscured, thereby achieving universal adaptability while managing system complexity through unified control.
Solution Approach 2:
The detection system achieves multi-functionality by incorporating both non-obscured and obscured face detection capabilities. The same system can handle various face conditions (obscured or non-obscured) using a unified processing framework that adapts to the input conditions.
Data Source
AI summary
A device and a method for calculating a swinging direction of a human face in an obscured human face image are provided. The method includes the following. An obscured human face image including a human face is captured. Non-obscured face detection technology is used to obtain a feature anchor point to be replaced in the obscured human face image, obscured face detection technology is used to obtain a plurality of candidate feature anchor points in the obscured human face image, and the plurality of candidate feature anchor points are used to determine an updated feature anchor point corresponding to the feature anchor point to be replaced. An adjustment operation is performed on a three-dimensional model to obtain an adjusted three-dimensional model. The updated feature anchor point and the adjusted three-dimensional model are used to calculate a swinging direction of the human face.