Researchers Develop AI Software for Auto-analysis of Hockey Game Videos

AI engineers and data scientists have developed a new AI software to automatically analyze videos of recorded hockey games. The technique involves automatically combining two existing Deep Learning (DL) techniques to identify players on the hockey rink / turf and analyze their performance using data science.

The Force that Developed AI Software for Hockey Game Video Analytics

The Vision and Image Processing (VIP) Lab created the latest AI software. The VIP Lab is a dedicated research group involved in AI Machine Learning innovations at the University of Waterloo. The VIP Lab is dedicated to understanding visual processes and finding solutions for the outstanding problems in visual processing and perception, as well as artificial intelligence, machine learning, and intelligent systems for a wide variety of applications. The research group is also a proud member of Partnership on AI.

Top Insights on AI ML Domain: 5G, IoT and AI: The Work of Art and Technology at Play in 2021

How Video Analysis Works with AI Machine Learning

The VIP Lab researchers built a comprehensive database of over 54000 different images extracted from the National Hockey League (NHL) Games. The AI algorithm was used to recognize the jersey number of the players based on multi-tasking learning. The ML algorithm involved data labeling with digit-wise representation to track players in the video, locate them on the ice rink, and then evaluating their performance throughout the game — all integrated into a single system of data analysis.

The video analysis of hockey games using AI software would be useful for journalists, trainers, and broadcasters

A spokesperson at the University of Waterloo stated: “As you can imagine, a person manually annotating video of a full hockey game of three periods would take hours. Machine-learning systems can produce data from videos in a matter of minutes.”

The same AI software could be used in other multi-player games such as soccer, cricket, and basketball. It could be also used to auto-tag passengers at an airport terminal or at passport offices.

SaaS and Analytics Updates: Lighting New York Partners with CQL for Salesforce Commerce Cloud Replatform

[To share your insights with us, please write to sghosh@martechseries.com]