PyTorch Implementation of Deep Colorization

Our Implementation is derived from the Original Paper 'Deep Koalarization: Image Colorization usingCNNs and Inception-Resnet-v2': https://arxiv.org/pdf/1712.03400.pdf

This approach has been tested on Coco-dataset with 70000 train images, 5000 validation images and 10000 test images

The Coco dataset can be downloaded on AWS or Google Colab using the below command line arguments:

wget -N images.cocodataset.org/zips/train2017.zip
wget -N images.cocodataset.org/zips/val2017.zip
wget -N images.cocodataset.org/zips/test2017.zip

Final Report

Final Results

Using Tensorboad on AWS

pip install tensorboard
run the tensorboard using 'tensorboard --logdir==runs"
Note the pytorch version has to be >=1.2