How to use Convolutional Neural Network to Build an AI Referee

Paper Title

How to use Convolutional Neural Network to Build an AI Referee

Authors

Rui Guo, Culver Academies, USA

Abstract

The purpose of this paper is to solve a significant problem in fencing: high-level referees are expensive, and low-level referees might not make correct and consistent calls. This research project compared a computer-based program to a referee's performance to solve the problem of inconsistency and frequency of errors. The paper has four sections, including background information, methodology, results, and a conclusion. The introduction includes the motivation and location for the research. The background consists of the basic saber fencing rules, defines machine learning, and describes the Convolution Neural Network tool. Next, the methodology section includes four critical subject areas human post estimation, the referee, training an Artificial Intelligence (AI) program, and testing the program's accuracy by examining fencing videos. Only two of the videos tested were judged incorrectly among all fifteen videos tested, and both had valid reasons. Therefore, the program was relatively accurate.

Keywords

Saber Fencing, Machine Learning, Convolutional Neural Network, Referee

Volume URL: https://airccse.com/adeij/vol3.html

Pdf URL: https://airccse.com/adeij/papers/3322adeij02.pdf

Here's where you can reach us : adeij.journal@yahoo.com or adeij.journal@airccse.com or adeij@aircconline.com

#saberfencing #machinelearning #convolutionalneuralnetworks #referee #engineering #callforpapers #researchpapers #cfp #researchers #phdstudent #researchScholar #journalpaper #submission #journalsubmission