Weak Identifier Reordering for Image Detection Speed
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current techniques for detecting objects, such as people, from images using edge extraction and statistical learning boosting are limited in speed and accuracy, as they rely on weak discriminators that do not efficiently terminate operations, leading to suboptimal performance.
Innovation Solution
An information processing device and method that calculates scores for positive and negative images using weak identifiers, realigns weak identifiers based on maximum scores or learning errors, and employs a genetic algorithm to optimize the arrangement of weak identifiers for improved detection speed and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If weak identifiers are processed in the learned order using statistical learning boosting, then the detection accuracy is maintained, but the processing speed is reduced due to inability to terminate operations early
Solution Approach 1:
The patent pre-calculates and stores the minimum score values for positive images before the actual detection process. This preliminary preparation enables the system to compare against pre-established thresholds during detection, allowing early termination when scores fall below these thresholds, thus improving processing speed without compromising detection accuracy
Solution Approach 2:
The patent dynamically adjusts the detection process by enabling early termination based on score comparisons. The system adaptively stops processing for negative images when their scores fall below the pre-calculated minimum scores of positive images, creating a dynamic detection flow that optimizes speed while maintaining accuracy
2Reliability
If the number of weak identifiers is increased to improve detection accuracy, then recognition reliability is enhanced, but the computational complexity and processing time increase
Solution Approach 1:
The system pre-computes and stores minimum score values for all positive images across all weak identifiers before detection. This preliminary action creates a reference framework that enables rapid comparison during detection, allowing the system to handle multiple weak identifiers efficiently without proportional increases in processing time
Solution Approach 2:
The patent extracts and stores only the critical minimum score values from the set of all possible scores. By taking out only the essential threshold information needed for comparison, the system reduces the computational burden while maintaining the ability to evaluate multiple weak identifiers for accurate detection
Data Source
AI summary
An information processing device includes a first calculation unit which calculates a score of each sample image including a positive image in which an object as an identification object is present and a negative image in which the object as the identification object is not present, for each weak identifier of an identifier including a plurality of weak identifiers, a second calculation unit which calculates the number of scores when the negative image is processed, which are scores less than a minimum score among scores when the positive image is processed; and an realignment unit which realigns the weak identifiers in order from a weak identifier in which the number calculated by the second calculation unit is a maximum.