This repository contains an implementation of a sign detection model using a combination of Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks. The model leverages the MediaPipe library for extracting keypoint information from video frames.
The goal of this project is to detect and recognize signs or gestures made by a person in front of a camera. The model combines the power of CNNs for spatial feature extraction with LSTMs for temporal modeling. Key point information is extracted using the MediaPipe library, providing a rich representation of hand movements.
- Python 3.x
- TensorFlow
- MediaPipe
- NumPy