Facial Expression Removal from 3D Images for recognition purposes

We present an encoder-decoder neural network to remove deformations caused by expressions from 3D face images. It receives a 3D face with or without expressions as input and outputs its neutral form. Our objective is not to obtain the most realistic results but to enhance the accuracy of 3D face recognition systems.

Implementation Details

Normalization

Network Model

Installation

Requeriments

• Python 3+
• C++ 11+
• PointCloud Library (PCL)
• Tensorflow for GPU (and its dependencies)
• OpenCV
• Scipy, Scikit-learn
• RAM enough to hold all your train images

Step-by-Step

1. Clone the repository
2. Verify the dependencies installation with install.sh
3. Execute a specific code (normalization, inference, etc...)

Usage

Normalize samples

Process images

Train with your own images

Contributing

Pull requests are welcome. Issues request too. Feel free to fork this project and modify whatever you need. Please, give the credits.

Cite this work as

License

GNU GLPv3.0