View Proposal

Proposer
William Yoo
Title
Integrating Large Language Models with Computer Vision for Real-Time Tennis Match Analysis and Outcome Prediction
Goal
In this project we aim to explore the capabilities of using Gen AI in the world of tennis.
Description
By utilising suitable open-sourced large language models for contextual understanding and content generation, incorporating with specific computer vision models for taking in visual input, such as videos of tennis matches, and turning them into data which the machine can read. Once the video has been provided the machine will be able to classify and generate tennis match features, such as type of shot, court surface, point outcome and more. This data will be useful for the machine to predict outcomes and make statistical analysis. This initiative aims to allow the machine to watch and review tennis matches to then make real time predictions of the match outcome down to the final score as well as producing a comprehensive statistical report representing the analysis of each player’s performance in the game. To do this, machine learning concepts has to be taken into account where the model will be fit with large amounts of open-source tennis match data and trained to predict target variables which will be the final score and winner of the match.
Resources
Zhang, A., Lipton, Z. C., Li, M. and Smola, A. J. (2023). Dive into Deep Learning. Cambridge University Press. https://D2L.ai (Key text: read Chapters 9, 10 ,11) Buhamra, N. and Groll, A. (2025). Statistical enhanced learning for modeling and prediction tennis matches at Grand Slam tournaments. Arxiv. https://arxiv.org/pdf/2502.01613 Dumovic, M. and Howarth, T. (2018). Tennis Match Predictions Using Neural Neworks. Stanford University CS230 Project Report. https://cs230.stanford.edu/projects_spring_2018/reports/8290687.pdf HumanAbilityAI. (2025, January 15). Pushing the limits of LLMs and GenAI: My Journey to an AI Tennis Coach App. https://humanabilityai.co.uk/pushing-the-limits-of-llms-and-genai-my-journey-to-an-ai-tennis-coach-app Lei, Y., Lin, A. and Cao, J. (2024). Rhythms of Victory: Predicting Professional Tennis Matches Using Machine Learning. IEEE Access, vol. 12, pp. 113608-113617. https://doi.org/10.1109/ACCESS.2024.3444031 Wong, M. R. and Ramos, L. (2018). Atari Tennis AI Agent. Stanford University CS221 Course Poster. https://web.stanford.edu/class/archive/cs/cs221/cs221.1192/2018/restricted/posters/marcrowo/poster.pdf
Background
Url
Difficulty Level
Moderate
Ethical Approval
None
Number Of Students
1
Supervisor
William Yoo
Keywords
tennis, large language models, transformers, ai agent
Degrees
Bachelor of Science in Statistical Data Science
Back to List