This code is based on face-parsing.PyTorch project.
This is the Bisenet network used for face parsing aextraction, you shouldn't change anything of it. It is provided by face-parsing.PyTorch project.
This is the main test execution code. You need to execute this code to get the parsing structure acquisition and joint output of face by calling bisnetv2.py.
Results are:
 |
background | skin | brow | eye | nose | mouth |
|---|---|---|---|---|---|---|
| Paring Output |  |
 |
 |
 |
 |
 |
| Joint Output |  |
 |
 |
 |
 |
 |
- python3
- numpy
- matplotlib
- scipy
- seaborn
- PIL
- python-opencv
- pytorch
These packages may not be included in Anaconda and require additional installation
sudo apt-get install python-opencv