Interlinked Convolutional Neural Network for Face Parsing

This is a pytorch implementation of the ICNN paper.

How to Run

Copy and unzip Helen dataset in "../data" (one level up from project directory)

Prepare facial parts (Only once) python3 extract_parts.py

Resize dataset (Only once) python3 resize_dataset.py

Create result folder mkdir res/

Train Stage1 python3 train_stage1.py

Train Stage2 python3 train_stage2.py

Modulate background python3 bg_modulate.py

Run end2end python3 end2end.py

Name		Name	Last commit message	Last commit date
Latest commit History 49 Commits
README.md		README.md
bg_modulate.py		bg_modulate.py
data		data
end2end.py		end2end.py
extract_parts.py		extract_parts.py
ircnn.png		ircnn.png
model.py		model.py
prepare_landmarks.py		prepare_landmarks.py
preprocess.py		preprocess.py
requirements.txt		requirements.txt
resize_dataset.py		resize_dataset.py
test_stage1.py		test_stage1.py
test_stage2.py		test_stage2.py
train_stage1.py		train_stage1.py
train_stage2.py		train_stage2.py
utils.py		utils.py