Fitness monitoring system using body pose estimation to count repetitions of exercises like push-ups, jumping jacks, curl-ups, and squats. Tracks yoga pose durations and include an alert for animal detection in the exercise area, ensuring safety and accuracy in real-time using YOLOv*8.
AI Zenith (AIZ) is a sophisticated machine learning project that combines exercise repetition counting with yoga pose detection using YOLOv8. The system's primary objective is to detect and count repetitions of common exercises like push-ups, jumping jacks, curl-ups, and squats by recognizing human body movements through Body Pose Estimation technology. Additionally, it classifies yoga poses, tracks the duration of each pose, and includes a safety feature that alerts users if a pet enters the exercise space to prevent accidents.
- Nay Myo Kyaw
- Khin Chaw Thazin
- Nay Lin Htun
- Min Khant Wunna
- Exercise Detection: Identifies and counts specific exercise repetitions.
- Yoga Pose Classification: Detects and classifies yoga poses while tracking the duration.
- Safety Alerts: Alerts users in real-time if a pet enters the workout area to prevent potential injuries.
This project uses a custom dataset comprising images tagged with various yoga poses and exercises. It also includes annotations for animal detection to trigger alerts during workouts.
- YOLOv8: Utilized for detecting human poses and classifying them.
- Body Pose Estimation: Employs pose estimation to determine the type of exercise and the execution of yoga poses.
To run this project locally, you will need to have Python and the necessary libraries installed. Clone the repository to view.