Welcome to the repository for a real-time, AI-powered hand gesture recognition system designed to bridge the communication gap for the deaf and mute community. This application leverages Computer Vision and Machine Learning to recognize hand gestures and interpret them in real time.
This project was developed as a major academic initiative (Sep 2023 – Dec 2023), focusing on inclusive technology and real-world impact. Built using Python, MediaPipe, and OpenCV, the system accurately detects and classifies hand gestures, displaying corresponding text-based communication in a user-friendly GUI.
- 🧠 AI-Powered Prediction: Integrated a custom-trained RandomForestClassifier model to recognize hand gestures using extracted landmarks.
- ✋ Precise Hand Detection: Leveraged MediaPipe Hands API for accurate real-time landmark tracking and gesture extraction.
- 🖥️ Real-Time Recognition: Seamless recognition pipeline using OpenCV, optimized for performance on standard hardware.
- 🪟 Accessible GUI: Built with Tkinter, offering an intuitive, easy-to-use interface suitable for diverse users.
- 🗂️ Modular Codebase: Includes clearly separated modules for data collection, training, processing, and inference.
- Programming Language: Python
- Libraries: OpenCV, MediaPipe, Tkinter, NumPy, Scikit-learn
- Model: RandomForestClassifier (sklearn)
- Development Tools: VS Code, GitHub, Jupyter Notebook
├── Application.py # Main application interface
├── Data_Collection.py # Module for collecting landmark data
├── Data_Preprocessing.py # Preprocessing raw data for ML
├── Data_Training.py # Training the classifier model
├── Sample_Model.ipynb # Jupyter notebook version of training
├── data.pickle # Trained model data
├── Requirements.txt # Required dependencies
├── data/ # Landmark data folder
- Capture hand landmarks in real-time using MediaPipe.
- Extract features and classify them using a trained Random Forest model.
- Display the predicted gesture as readable text in a GUI.
- Practical understanding of real-time computer vision systems.
- Experience in building end-to-end machine learning applications.
- Improved skills in Python, GUI design, and model training.
- Developed empathy-driven tech for assistive communication.
Hi 👋, I'm SYED IBRAHIM A – a recent B.Tech CSE graduate with a specialization in AI (IBM Certified). I'm passionate about using technology for good, especially in fields like accessibility, healthcare, and education.
- 💼 Open to: Python Developer, AI/ML Developer, Software Engineer, QA Engineer
- 🌐 Portfolio: https://www.linkedin.com/in/ibrahimcreator/
- 📍 Location: India (Open to Remote / Hybrid / On-site opportunities)
This project is licensed under the MIT License.
If this project inspires you or aligns with your goals, feel free to star ⭐ it and connect with me!