Action Recognition, Tracking, and Optimization Analysis of Training Process Based on the Support Vector Regression...

Area of Science:

• Computer Science
• Machine Learning
• Biomedical Engineering

Background:

• Human action recognition is crucial for applications like surveillance and human-computer interaction.
• Existing methods often face challenges in accuracy and efficiency.
• Support Vector Machines (SVMs) are powerful tools for classification tasks.

Purpose of the Study:

• To propose an optimized Support Vector Machine (SVM) method for enhanced human action recognition.
• To improve the accuracy and efficiency of action recognition training processes.
• To integrate confidence levels for more robust recognition results.

Main Methods:

• Developed a novel human action recognition method utilizing an optimized Support Vector Machine (SVM).
• Implemented a Directed Acyclic Graph (DAG) SVM strategy, improved based on classifier accuracy.
• Incorporated confidence levels with recognition outputs for refined result processing.

Main Results:

• Achieved a high recognition rate of 98.7% for human actions using the proposed SVM optimization.
• Demonstrated significant improvements in both accuracy and efficiency of the action recognition process.
• Validated the effectiveness of using confidence levels to process recognition outcomes.

Conclusions:

• The proposed SVM optimization method is highly effective for human action recognition.
• This approach significantly enhances the accuracy and efficiency of recognizing human body actions.
• The integration of confidence levels further refines the action recognition system.