Code for Unsupervised Learning of Compositional Energy Concepts

This is the pytorch code for the paper Unsupervised Learning of Compositional Energy Concepts.

Demo

Please download a pretrained model at this link and then execute the following code to test a pretrained CelebA-HQ 128×128 COMET model

python demo.py im_path=im0.png

Please utilize the following command to run global factor decomposition on CelebA-HQ (or other datasets)

python train.py --exp=celebahq --batch_size=12 --gpus=1 --cuda --train --dataset=celebahq --step_lr=500.0

You may further run the code on high-resolution 128×128 images below

python train.py --exp=celebahq_128

 
 

 
To finish reading, please visit source site