University of Twente Student Theses

Login

Enhancing Live Commentary Generation in Soccer Video Games through Event Prediction with Machine Learning Methods

Kościołek, Jakub (2024) Enhancing Live Commentary Generation in Soccer Video Games through Event Prediction with Machine Learning Methods.

[img] PDF
502kB
Abstract:This paper addresses the challenge of enhancing live commentary generation in soccer video games through the prediction of in-game events using machine learning methods. Traditional prerecorded commentary systems fail to adapt dynamically to the evolving narrative of the game, often resulting in repetitive commentary. Attempts to solve this problem by integration of Large Language Models and Text-to-Speech technology generate additional challenges connected with overhead that is needed for those technologies resulting in noticeable delay. This research focuses on mitigating these issues by leveraging Support Vector Machine (SVM) and Artificial Neural Network (ANN) models to predict events such as Goal Kicks, Free Kicks, Corners, and Throw Ins, seconds before they occur. Utilizing data from the Google Football Environment, we trained and tested these models, examining their performance. Our findings indicate that while there is potential in such a methodology, further improvements need to be made to ensure that the model is working well in real-world scenarios. This study provides a foundation for future improvements in real-time commentary generation, emphasizing the potential of machine learning to minimize delays in generated commentary, thereby enhancing the immersive experience of soccer video games.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Computer Science BSc (56964)
Link to this item:https://purl.utwente.nl/essays/100983
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page