Exploring the Application of Computer Vision in Elite Sports – The Value Proposition and Integration Challenges

by Varun Kodathala | at Minnebar 17

Technology has a profound impact on many industries, including the sports. Artificial Intelligence in particular has completely disrupted the sports industry, enabling the precise analysis of the performances of players. SportsVision is an advanced sports analytics platform powered by Artificial Intelligence, Deep Learning and Computer Vision for both off and on the field analytics that caters to the needs of everyone involved in sports management. It enables coaches and managers to analyze and enhance the performances of players with vision-centric insights like never before. Once the videos are uploaded onto the SportsVision platform, the deep learning algorithms provide multiple insights such as automated actions, event tagging, pose analysis, player tracker, and rating application as needed.

Detailed Outline of Presentation: Introducing SportsVision: • Problem Statement • Our Solution • Our Evolution Artificial Intelligence for the Win: • Assessments and Training Evaluation • Performance Analysis • Tactical Analysis The SportsVision Paradigm: • Anthropometric Analysis • Computer Vision based Bio-Mechanical Analysis • AI based Individual player performance statistics • Individual Strategic Analysis • Computer Vision driven SWOT Analysis for both home and opponent team tactics Key Challenges of AI usage in Elite Sports: • Data Collection • Connecting AI and Sports Communities • Keeping control on decision making • Interpretability of AI results • Robust Modelling Platform Illustration of AI based Stats in: • Diving • Boxing • Taekwondo Summary and Conclusions of AI Usage in Elite Sports: • Risk • Potential • Future and Ease of use in sports

All levels

Varun Kodathala

Varun Kodathala is an Chief Architect/ AI engineer from Vizworld Inc., and he masters the tech stack around computer vision and its integrations. He holds a bachelor's degree with majors in Electronics and Communication Engineering and specializes in digital image processing and electronics. He is a recipient of two university gold medals and a distinguished student award from the Malaysian Journal during his bachelor's. In prominent journals, he published about 23 research articles in discrete electronics, image processing, and computer vision. One of his research articles was indexed in The SAO/NASA Astrophysics Data System, and global researchers cited his research frameworks.