Skeleton Information Determination via Joint Confidence Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional skeletal information determination methods face challenges in accurately analyzing the state of a photographed person due to errors in joint point recognition and estimation, leading to manual cleansing processes that are time-consuming and hinder real-time processing.
Innovation Solution
A skeletal information threshold setting apparatus that sets confidence thresholds for joints based on important and non-important joints, using linear interpolation to determine the quality of skeletal estimation results, allowing for automatic determination of usable skeletal information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual cleansing of skeletal estimation results is performed to ensure accuracy, then state analysis accuracy is improved, but processing time increases and real-time processing becomes difficult
Solution Approach 1:
The system performs preliminary quality assessment of skeletal estimation results by comparing confidence scores against thresholds before state analysis is conducted. This preliminary filtering action identifies and removes low-quality estimations in advance, ensuring that only high-confidence results proceed to state analysis, thereby maintaining accuracy without requiring time-consuming manual intervention later in the process.
Solution Approach 2:
The system implements automated quality determination through confidence score comparison, where the skeletal information determination device itself evaluates and filters its own estimation results without external manual intervention. The device autonomously identifies unreliable joint point estimations by checking whether confidence scores meet predetermined thresholds, enabling self-service quality control that eliminates the need for manual cleansing while maintaining processing speed.
2Device complexity
If all joint points are assigned the same confidence threshold, then the determination process is simplified, but accuracy decreases due to varying importance of different joints
Solution Approach 1:
The system assigns different confidence thresholds to different joint points based on their relative importance and reliability characteristics. Critical joints such as those in the upper body receive higher thresholds requiring greater confidence, while less critical joints receive lower thresholds. This localized quality approach ensures that the skeletal estimation quality meets appropriate standards for each specific joint's functional importance, optimizing overall state analysis accuracy without uniform constraints.
Solution Approach 2:
The determination process is segmented into multiple stages: first, joint points are categorized by importance (e.g., upper body vs. lower body joints); second, different threshold criteria are applied to each category; third, quality determination is performed hierarchically. This segmentation allows the system to manage complexity by breaking down the overall determination process into manageable segments with tailored threshold requirements, rather than applying a single complex threshold to all joints simultaneously.
Data Source
AI summary
A skeletal information threshold setting apparatus includes: a joint information input unit that accepts an input of important joints among joints of a subject and a confidence threshold for the important joints; and a threshold setting unit that acquires a confidence threshold for each of multiple joints of the subject, including the important joints, based on the important joints and the confidence threshold for the important joints that were input, and sets the acquired confidence thresholds for the joints as thresholds to be used in making a determination regarding a skeletal estimation result for the subject.