Skip to content

0xanwar/LipReader

Repository files navigation

Lip Reader

Overview

This project focuses on implementing Lip Reading using 3D convolution with TensorFlow, Keras, and OpenCV. The system is designed to recognize English language lip movements from video data.

Table of Contents

Required Libraries

Before running the Lip Reader project, make sure to install the required libraries. You can install them using the following command:

pip install opencv-python tensorflow numpy matplotlib imageio

Required Libraries

pip install opencv-python tensorflow numpy matplotlib imageio

OpenCV (cv2): Used for image and video processing. TensorFlow: Deep learning framework for building and training models. NumPy: Library for numerical operations, particularly useful for handling arrays and matrices. Matplotlib: Used for creating visualizations and plots. ImageIO: Library for reading and writing image data.

Installation

1.Clone the repository:

git clone https://github.com/ahmedanwar123/LipReader.git
cd LipReader

2.Install dependencies:

pip install -r requirements.txt

Usage

1.Open the Jupyter Notebook:

jupyter notebook LipReader.ipynb

Architecture

The lip reading model architecture utilizes a 3D convolutional neural network (CNN) for feature extraction from video frames. The model is trained on a dataset of labeled lip movement sequences.

Data Preparation

Preparing training data by organizing video sequences and corresponding labels. The dataset should include English language speakers showcasing various lip movements.

Training

To train the model, follow these steps:

Organize your dataset. Run the training script in the notebook. Evaluation Evaluate the trained model on a separate test set using the provided evaluation script in the notebook.

Streamlit App

streamlit run streamlitapp.py

Future Work

Language Expansion: Include support for Arabic language lip reading. Enhanced Model: Experiment with more advanced 3D convolutional architectures. Real-time Processing: Implement real-time lip reading capabilities. Multilingual Support: Extend language support to other languages.# Lip Reader

Overview

This project focuses on implementing Lip Reading using 3D convolution with TensorFlow, Keras, and OpenCV. The system is designed to recognize English language lip movements from video data.

Table of Contents

Required Libraries

Before running the Lip Reader project, make sure to install the required libraries. You can install them using the following command:

pip install opencv-python tensorflow numpy matplotlib imageio

Required Libraries

pip install opencv-python tensorflow numpy matplotlib imageio

OpenCV (cv2): Used for image and video processing. TensorFlow: Deep learning framework for building and training models. NumPy: Library for numerical operations, particularly useful for handling arrays and matrices. Matplotlib: Used for creating visualizations and plots. ImageIO: Library for reading and writing image data.

Installation

1.Clone the repository:

git clone https://github.com/ahmedanwar123/LipReader.git
cd LipReader

2.Install dependencies:

pip install -r requirements.txt

Usage

1.Open the Jupyter Notebook:

jupyter notebook LipReader.ipynb

Architecture

The lip reading model architecture utilizes a 3D convolutional neural network (CNN) for feature extraction from video frames. The model is trained on a dataset of labeled lip movement sequences.

Data Preparation

Prepare your training data by organizing video sequences and corresponding labels. The dataset should include English language speakers showcasing various lip movements.

Training

To train the model, follow these steps:

Organize your dataset. Run the training script in the notebook. Evaluation Evaluate the trained model on a separate test set using the provided evaluation script in the notebook.

Streamlit App

streamlit run streamlitapp.py

Future Work

  • Language Expansion: Include support for Arabic language lip reading.
  • Enhanced Model: Experiment with more advanced 3D convolutional architectures.
  • Real-time Processing: Implement real-time lip reading capabilities.
  • Multilingual Support: Extend language support to other languages.

About

A project aimed to read Lips in English and add Arabic Lip reading in future

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors