Logo Lanfrica

ZekariyasT/ESLTS

Domaine:

natural language processing

Type de record:

software
Créateur:
Zek
Hôte:
Ethiopian Sign Language Translation System # Real-Time Sign Language Detection and Translation This project aims to detect and translate Ethiopian sign language gestures in real-time using LSTM Deep learning neural network. The application, built as a desktop app using Python Tkinter, captures live video from the camera, processes it to detect hand gestures using the MediaPipe library, and translates them into Amharic text. ## Features - Real-time sign language detection and translation - Desktop application built with Python Tkinter - Uses a pre-trained deep learning model for gesture recognition - Simple and intuitive user interface - Works with any standard webcam ## Requirements - Python 3.x - Tkinter - OpenCV - NumPy - Pillow (PIL) - TensorFlow - MediaPipe ## Installation 1. Clone this repository to your local machine. 2. Install the required Python packages by running `pip install -r requirements.txt`. 3. Run the application using `python main.py`. ## Usage - Press the "Start Camera" button to begin capturing video from the webcam. - Perform sign language gestures in front of the camera. - The detected gestures will be translated into text and displayed on the screen. - Press the "Stop Camera" button to stop capturing video.