Locked in Lacrosse - Player Performance Analysis App

Summary

Locked In Lacrosse - a lacrosse training provider based in New Jersey, sought Folio3's expertise to develop an AI-powered performance analysis app. The AI app processed training videos and extracted insights based on form and pose to determine optimal and tailored training at scale.

Customer

Founded by New Jersey college lacrosse alumni - Locked In Lacrosse helps aspiring players of all ages through specialized cross-training programs, incorporating advanced techniques to garage and improve player’s performances. As part of their mission to provide top-notch training, they continuously strive to improve the quality and techniques of their training.

Understanding the Challenge

It was quite time-consuming to assess each player's performance manually. Locked in Lacrosse needed an AI solution that could improve the training quality, assess the player's form and technique, identify areas for improvement, and prevent injuries—improving the overall training experience.

Solution

Folio3 AI built an AI-powered web application based on the activity detection model that enables trainers to analyze lacrosse players' performance through video analysis. Using advanced pose estimation techniques, the application assesses specific movements of players, generates an output video with pose markings, and calculates metrics/results based on those movements.

• Pose Estimation: The system will utilize state-of-the-art pose estimation algorithms to analyze player form and detect key biomechanical markers to determine optimal pose while shooting, pitching, etc
• Injury Prevention: Assess the player's form and technique, to identify optimal pose and prevent injuries.
• Player Performance Analysis :Detect and overlay pose information onto video frames, highlighting key body points and providing visual cues for player performance analysis and user interaction.
• Calculation and Results: Perform real-time calculations on the pose estimation data and provide results on player performance based on biomechanical markers

Result

The web application improved players' training experience with over 90% accuracy and led to significant performance changes.