ViewU is an automated tool, that can enable user to have of view of him/her on the products he/she purchase like dresses, accessories, jewelry etc.This is implemented using GANs
To train the model, we used LOOKBOOK dataset, resize images to 64*64. Prepare the dataset using pytorch-GAN/tool/prepare_data.py
.
Then goto src dir and run
python3 train.py
-
Install visdom.
-
Start the visdom server with
python3 -m visdom.server 5274
-
Open this URL in your browser:
http://localhost:5274
You will see the loss curve as well as the image examples. -
Development in Progress