✋ MULTI GESTURE DETECTOR 🤖

This project is a real-time hand gesture recognition system using MediaPipe, OpenCV, and a TensorFlow/Keras deep learning model. It allows users to perform various hand gestures that are detected live via webcam.

---

🚀 Features

• ✅ Real-time hand gesture detection using webcam
• 🧠 Custom-trained model using TensorFlow/Keras
• 📷 MediaPipe for accurate hand landmark tracking
• 🎯 Confidence threshold and cooldown logic for stable predictions
• 🔁 Support for multiple gestures including:
  • Hello 👋
  • Fist ✊
  • Thumbs Up 👍
  • Thumbs Down 👎
  • Peace ✌️
  • Stop ✋
  • Okay 👌
  • Rock 🤘
  • Call Me 🤙

---

🛠️ Tech Stack

• Python
• OpenCV
• MediaPipe
• TensorFlow / Keras
• NumPy

---

📁 Directory Structure

MULTI-GESTURE-DETECTOR/
│
├── model.keras                 # Trained gesture classification model
├── gestures/                  # Folder containing .csv data for each gesture
├── gestures.txt               # List of gesture labels
├── collect_data.py            # Script to record gesture data
├── train_model.py             # Script to train the model
├── predict_gesture.py         # Main script for real-time prediction
└── README.md                  # This file

---

🧪 How to Run

1. Install Dependencies

   pip install opencv-python mediapipe numpy tensorflow

2. Run Real-Time Prediction

   python predict_gesture.py

3. Press q to quit the live window

---

🧠 Model Training

To train your own model:

1. Record data using collect_data.py
2. Train the model using train_model.py
3. Ensure the model is saved as model.keras

---

🤩 Future Enhancements

• 🔥 Two-hand gesture detection
• 🌐 Web-based UI with Flask
• 💾 Save gesture history/log
• 🧠 Integrate voice feedback or system control

---

📜 License

This project is licensed under the MIT License.

---

👤 Author

Vansh Agrawal