raahatg21 / Eyes-Lips-Segmentation

Eyes and Lips Segmentation in a Face

Python program which inputs an image of a face, and creates a mask around the eyes and lips of the face.

Overview

The network architecture used here is based on BiSeNet [1] and ResNet. The dataset used is CelebAMask-HQ Dataset [2].

How to run

Prerequisties

• Python 3
• PyTorch
• Scikit-Image
• OpenCV
• Pillow

Steps

• Download the pre-trained model from here. Put the .pth file in cp folder.
• Run eyes-and-lips.py with the path of the image as argument. For example:

python eyes-and-lips.py --img-path images/img1.jpg

Results

Input	Mask	Output

Sources

• [1] BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation
• [2] CelebAMask-HQ Dataset
• [3] https://github.com/zllrunning/face-parsing.PyTorch