Swoleboi: Pose Detection and Repetition Counter

Overview

Swoleboi is a real-time pose detection and repetition counting application designed to monitor specific exercises like deadlifts. Using a webcam, it analyzes body posture and movements, providing feedback on the stage of the exercise and counting repetitions.

Features

• Real-time Pose Detection: Utilizes MediaPipe to capture and analyze body landmarks.
• Repetition Counting: Counts repetitions based on the detected stages of exercise.
• Probability Display: Shows the probability of the detected pose class.
• Custom UI: Built with Tkinter and CustomTkinter for a dark-themed, custom interface.

Requirements

• Python 3.x
• Tkinter
• CustomTkinter
• Pandas
• NumPy
• OpenCV
• Pillow
• Mediapipe

Installation

1. Clone the repository:

   git clone https://github.com/NitinRwt/MLapp.git
   cd MLapp

2. Install the required packages:

   pip install -r requirements.txt

3. Place the model file:

   • Ensure deadlift.pkl is in the project directory. This file contains the trained model for pose classification.

Usage

1. Run the application:

   python swoleboi.py

2. Interface Details:

   • STAGE: Shows the current stage of the exercise (e.g., "up" or "down").
   • REPS: Displays the count of repetitions completed.
   • PROB: Indicates the probability of the current pose classification.

3. Reset Counter:

   • Click the "RESET" button to reset the repetitions counter.

Configuration

• Webcam Settings:

  • The default camera index is set to 3. Adjust cv2.VideoCapture(3) to your correct camera index if needed.

• Pose Confidence:

  • Adjust the min_tracking_confidence and min_detection_confidence in mp_pose.Pose() for different detection sensitivity.

Troubleshooting

• Camera Issues:

  • Ensure the correct camera index is used.
  • Verify that the camera is not being accessed by another application.

• Model Errors:

  • Ensure deadlift.pkl is in the correct location and is not corrupted.

Contributing

Feel free to fork the repository and submit pull requests. For major changes, please open an issue first to discuss what you would like to change.

License

This project is open-sourced under the MIT License. See the LICENSE file for more details.

Acknowledgments

• Thanks to the MediaPipe team for their pose detection library.
• Inspired by fitness applications that utilize AI for exercise monitoring.