Skill Assessment Using Confidence Measures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing human activity recognition systems fail to assess the skill level involved in performing detected activities, as they focus on activity type recognition without evaluating quality, and domain-specific methods are limited to their original application domains.
Innovation Solution
A computationally inexpensive skill assessment method using machine learning models that classify activity types and skill levels based on confidence measures, employing feature vectors that are smaller than traditional methods, allowing for real-time operation and reduced power consumption, and applicable across various domains without requiring domain-specific features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex hierarchical analysis frameworks are used for skill assessment, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts only the essential confidence measure from activity recognition outputs, discarding complex hierarchical analysis frameworks. By taking out only the necessary confidence information and using it directly for skill assessment through simple statistical methods, the system achieves accurate skill evaluation without the computational overhead of complex hierarchical models.
2Measurement precision
If domain-specific skill assessment methods are used, then measurement precision is improved for that domain, but adaptability deteriorates
Solution Approach 1:
The patent creates a universal skill assessment framework that works across multiple domains by using domain-independent confidence measures from activity recognition systems. The method applies the same statistical analysis approach regardless of the specific activity domain, enabling the system to assess skill in sports, healthcare, and other domains without requiring domain-specific customization.
3Measurement precision
If traditional skill assessment methods are used, then measurement precision is improved, but use of energy increases
Solution Approach 1:
The patent uses lightweight, computationally inexpensive statistical methods instead of heavy machine learning models. By employing simple probability distributions and confidence measure analysis that require minimal computational resources, the system achieves accurate skill assessment while consuming significantly less energy, making it suitable for battery-powered wearable devices.
4Measurement precision
If complex feature vectors are used in classification, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts only the confidence measure from activity recognition outputs, discarding complex feature vectors. By taking out only the essential confidence information and using simple statistical methods to analyze it, the system achieves accurate skill classification without the computational burden of processing high-dimensional feature vectors.
Data Source
AI summary
According to an embodiment there is provided a method of skill classification comprising receiving data indicative of an activity performed by a person, classifying the type or types of activity performed by the person based on the received data, wherein classifying provides an indication of an activity type or activity types as well as an indication of the confidence that an activity has been classified correctly and classifying a skill level associated with a classified activity or classified activities on the basis of the indication of confidence.