Face Detection Prioritizing Inclination Probabilities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional face detection methods require significant processing power and time due to the need to detect faces at various inclinations, especially in non-uniform photography conditions, which increases the complexity of the detection process.
Innovation Solution
A method that determines the relative probabilities of face inclinations in input images and sets the order of detection based on these probabilities, optimizing the detection process by prioritizing the most likely inclinations first, particularly distinguishing between continuous and video imaging modes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If face detection is performed by repeating detection processes while varying face inclinations, then faces with different inclinations can be detected, but the amount of processing increases significantly
Solution Approach 1:
The patent applies preliminary action by performing a rough detection process before the detailed detection process. The rough detection identifies candidate regions and face inclinations, allowing the subsequent detailed detection to focus only on these candidates rather than processing the entire image at multiple inclinations, thus reducing overall processing complexity while maintaining detection coverage
Solution Approach 2:
The detection process is segmented into two distinct stages: rough detection and detailed detection. The rough detection stage performs quick scanning to identify potential face regions and their inclinations, while the detailed detection stage performs comprehensive analysis only on these identified candidates. This segmentation reduces the total processing load by avoiding redundant computations in non-face regions
2Adaptability or versatility
If detection processes are repeated while varying inclinations, then faces with unknown inclinations can be detected, but detection time increases
Solution Approach 1:
The rough detection process performs preliminary identification of face candidates and their inclinations quickly, allowing the system to adapt the detailed detection to match the detected inclination rather than trying all possible inclinations. This preliminary action significantly reduces detection time while maintaining the ability to handle unknown face inclinations
Solution Approach 2:
The patent introduces dynamic adaptation where the detection parameters (particularly inclination) are adjusted based on results from the rough detection stage. Instead of using a fixed set of predetermined inclinations, the system dynamically selects the inclination to use in detailed detection based on what was detected in the rough stage, optimizing detection time for each specific image
3Device complexity
If a single detection process is used, then processing is simplified, but faces with different inclinations cannot be detected
Solution Approach 1:
The detection process is divided into rough detection and detailed detection stages, where rough detection handles multiple inclinations quickly to identify candidates, and detailed detection focuses on a single inclination for each candidate. This segmentation maintains relative simplicity while achieving multi-inclination detection capability
Solution Approach 2:
The rough detection performs preliminary identification of face candidates and their inclinations, enabling the detailed detection to be customized for each candidate's specific inclination. This preliminary action allows the system to maintain simple processing for the majority of cases while adapting to handle different inclinations when necessary
Data Source
AI summary
An apparatus detects a predetermined number of facial images from detection target images. An inclination order setting means utilizes the correlative relationships among correlative data obtained by a correlative data obtaining means and the inclinations of faces that appear in input images, to determine the relative value of the probability that faces of a predetermined inclination will appear in the input images. The inclination order setting means sets the order of inclinations of faces to be detected such that faces are detected in order of inclinations according to the relative values of the probabilities, based on the correlative data that they will appear. A face detecting means detects faces within the input images while varying the inclinations of faces to be detected according to the set order.