Analysis of Sports Video Intelligent Classification Technology Based on Neural Network Algorithm and Transfer...

Area of Science:

• Computer Science
• Artificial Intelligence
• Digital Media

Background:

• Explosive growth in digital content necessitates efficient archiving.
• Accurate sports video classification is crucial for digital content management.

Purpose of the Study:

• To develop an advanced deep learning model for sports video classification.
• To improve the accuracy and efficiency of sports video categorization.

Main Methods:

• Utilized deep neural network (DNN) and convolutional neural network (CNN) algorithms.
• Implemented transfer learning with the maximum mean difference (MMD) algorithm.
• Introduced block brightness comparison coding (BICC) and block color histograms for feature extraction.

Main Results:

• The proposed deep learning coding model achieved superior performance in sports video classification.
• Demonstrated significant improvements in classification accuracy across various sports video types.
• Outperformed existing sports video classification methods.

Conclusions:

• The developed method offers a highly effective solution for sports video classification.
• Enhances digital content archiving through accurate and efficient video categorization.
• Represents a significant advancement in deep learning applications for sports media analysis.