🤟 Sign Language Detector

A real-time hand gesture recognition system that converts sign language into readable text or speech using computer vision and machine learning.

---

🧠 Features

• 📷 Collects custom hand gesture images using webcam
• 🧠 Trains a custom ML model using Random Forest
• 💡 Real-time detection and classification of hand signs
• 🔊 Optional text-to-speech feedback for recognized signs

---

💻 Tech Stack

• Languages & Libraries: Python, OpenCV, NumPy, scikit-learn, Tkinter
• ML Model: Random Forest Classifier
• Others: pickle, os, cv2, time

---

📁 Project Structure

sign-language-detector/
├── collect_imgs.py # Capture and store gesture images
├── create_dataset.py # Preprocess images and create labeled data
├── train_classifier.py # Train ML model and save as model.p
├── inference_classifier.py # Use webcam to detect gestures live
├── requirements.txt # Python dependencies
├── model.p # Trained classifier (small enough to include)
├── data.pickle # Labeled dataset (optional for upload)
├── README.md # This file

---

⚠️ Notes Before You Start

• Python 3.10 is required. This project will NOT work on Python 3.11 or 3.13 due to mediapipe compatibility issues.
• You need a working webcam.
• The project assumes a trained model.p file already exists. If you're cloning this repo fresh, it will run without re-training.

---

🔧 Setup Instructions

1️⃣ Install Python 3.10 👉 Download it here: Python 3.10 - https://www.python.org/downloads/release/python-3100/ ✅ Make sure to check "Add to PATH" during installation.

2️⃣ Clone the Repository git clone https://github.com/your-username/sign-language-detector-python.git cd sign-language-detector-python python collect_imgs.py

3️⃣ Create a Virtual Environment py -3.10 -m venv venv .\venv\Scripts\activate

4️⃣ Install Dependencies pip install opencv-python mediapipe==0.10.9 scikit-learn

Or install from requirements.txt: pip install -r requirements.txt

5️⃣ Run the Project python inference_classifier.py

The webcam will open.

Show a trained hand gesture (A, B, C, etc.)

Prediction will appear in the console.

🗂 Dataset Note For privacy and size reasons, the data/ folder (with training images) is not included.

If you'd like to retrain the model: python collect_imgs.py → Collect new images python train_classifier.py → Train model.p again

✅ Follow the same class labeling and folder structure as the original.

---

👩‍💻 Author

Rasagna Rayasam
Aspiring AI/ML Developer | Passionate about building tech that helps people
📫 Connect on GitHub

---

Thanks for checking out this project! ⭐ If you liked it, drop a star on the repo to support my work.