This open-ended project is for focusing on advanced sports analytics. The goal of this project is to detect lines in the fields and create visualizations from a given American football game video, to determine team influence of player behavior, etc.
We fully discretize DreamerV2 and evaluate the discrete representation learned by the model.
Various model-based methods for extracting interpretable insights from discrete time series.
Collection of capstone projects for final year CISC students
Dialogue Agents and Learning Constraints for SAT/CSPs