🤟 ISL Recognition - Indian Sign Language Detection Using CNN

This project is a real-time Indian Sign Language (ISL) recognition system using a Convolutional Neural Network (CNN) and OpenCV. It processes video input from your webcam and predicts hand gestures corresponding to ISL characters or words.

🧠 Model

The model (model.h5) is trained on grayscale images of ISL hand signs resized to 64x64 pixels. It outputs the predicted label using softmax classification.

🗂 Dataset

This project expects a folder named dataset/, where each subfolder is named after a sign class (e.g., A, B, Hello) and contains training images for that gesture.

dataset/
├── A/
├── B/
├── Hello/
...

🚀 Getting Started

✅ Prerequisites

• Python 3.x
• pip

📦 Install Dependencies

pip install numpy opencv-python keras tensorflow

You may need to install h5py as well: pip install h5py

🏁 Run the Project

python isl_recognition.py

Press q to quit the recognition window.

---

🔍 Project Structure

├── isl_recognition.py    # Main script for webcam-based recognition
├── model.h5              # Trained CNN model
├── dataset/              # Dataset used for training (not included)
└── README.md             # Project documentation

---

🔧 Future Improvements

• ✅ Add FPS counter and confidence scores.
• 🔲 Include gesture region detection for better accuracy.
• 🔲 Add training script and dataset download instructions.
• 🔲 Extend to full words or sentences using sequence models.

---

🤝 Contributing

Feel free to fork the repository, open issues, or submit pull requests.

---

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

🙏 Acknowledgements

• Keras
• OpenCV
• Public ISL datasets used for training (not distributed with this repo)