Hey there, I’m Marjan Shahi! I’m currently a Ph.D. student in the VIP’s sports analytics group, focusing on machine vision and deep learning. My research is all about developing a methodology for analyzing hockey player performance based on their actions’ value.
One of my primary areas of interest is player action recognition. By using advanced computer vision techniques and machine learning algorithms, I aim to identify and analyze specific player actions on the ice. This could be used in player evaluation techniques as input to analyze their performance and provide coaches and teams with valuable insights into player performance, such as identifying areas for improvement or optimizing their game strategies.
I’m really excited about the work I’ve been doing on goalie pose estimation, too. We’re working on predicting goalie, net, and equipment poses with high accuracy, which will be super valuable for improving goalie action recognition and performance evaluation.
Another area of research I’m interested in, and I’ve worked on during my Master’s is reinforcement learning. I want to improve the generalization ability of RL agents so that they can adapt to new environments and situations more effectively.
Before starting my Ph.D., I worked in the industry doing business analytics for a great company. It was a fantastic experience, and it’s given me a unique perspective on the intersection of academia and industry.
Overall, I’m really passionate about using data and technology to drive our understanding of the world. I’m excited to keep making significant contributions to the field of sports analytics and beyond!
Pose Estimation, Action Recognition