Abstract: Virtual assistants have become an integral part of our daily lives, significantly influencing how we perform various activities. With the rising prominence of AI, we aim to explore this emerging field through our project, which focuses on AI-based workout assistants. Introducing Workout assistant, our application is designed to detect users' exercise poses, count repetitions for specified exercises, and provide personalized recommendations to enhance their form. By utilizing MediaPipe, the application accurately identifies a person's pose and analyzes the pose's geometry using both dataset information and real-time video. This analysis enables Workout assistant to precisely count repetitions for specific exercises and offer detailed guidance to users on improving their exercise technique. Our goal is to leverage AI technology to create a comprehensive virtual workout assistant that empowers users to track their progress, ensure proper form, and gain valuable insights for their fitness journey.
Keywords: AI, Virtual assistant, CNN, workout assistant, Pose estimation, Blazepose, OpenCV.
|
DOI:
10.17148/IJARCCE.2023.12639
