Empowering communication through seamless gesture recognition.
no-lingo is a cutting-edge developer tool that empowers users to harness gesture recognition for real-time communication and image classification.
This project aims to simplify the integration of gesture-based interactions into applications. The core features include:
- ✋ Gesture Recognition: Processes image contours to translate gestures into text, enhancing accessibility.
- 📊 SVM Model Training: Leverages Histogram of Oriented Gradients (HOG) for effective image classification.
- 🎥 Real-time Sign Language Recognition: Uses live video to detect and translate hand gestures instantly.
- 📸 Image Dataset Creation: Captures and processes images to build robust, custom datasets.
- 🖌️ Skin Detection: Improves recognition by isolating hand regions using skin tone detection.
Ensure the following are installed:
- Python (recommended: 3.8+)
- Conda (for environment management)
Build no-lingo from source and install dependencies.
-
Clone the repository:
git clone https://github.com/shakirth-anisha/no-lingo
-
Navigate to the project directory:
cd no-lingo -
Create and activate the environment:
conda env create -f conda.yml
conda activate no-lingoTo run the application:
conda activate no-lingo
python ASL.pyFor Raspberry Pi 5 + Camera Module 3, you can use Picamera2 with the --raspi flag:
sudo apt install python3-picamera2
python ASL.py --raspi
python creatingDataset.py --raspi
python bounded.py --raspi