Skip to main content

GAN Based colouring model

Project description

ColourGAN

Simple PyTorch based GAN for colouring black and white images.

Installation

The code is available as python package.
Run following command for installing colourgan with all it's dependencies.

pip install colourgan

Additionally for trying out streamlit interface, please install streamlit using following command

pip install streamlit

You can also install it from this repository.

git clone https://github.com/narenderkumarnain/ColourGAN.git
cd ColourGAN
python setup.py install

Usage

Inference

For trying our pretrained model, please download the Generator Weight File
Now use these lines of code for inference on sample image.

#imports
import cv2
from colourgan.model import ColourGAN
from colourgan.config import get_cfg

# preparing the config
cfg = get_cfg()
# adding path to generator weights
cfg.initial_weights_generator = 'checkpoint_ep99_gen.pt'

# loading the model
model = ColourGAN(cfg , inference=True)

# loading the image
img = cv2.imread('sample.png')

# running inference
res = model.inference(img)

# saving result image
cv2.imwrite('sample_res.png' , res)

Streamlit interface

For trying out the streamlit interface, create a file app.py with following code.

# import
from colourgan.streamlit import app

# path to weights file
path_to_generator_weights = 'checkpoint_ep99_gen.pt'

if __name__ == '__main__':
    app(path_to_generator_weights)

We expect you have already installed streamlit.
Use followind code to run the application

streamlit run app.py

Training

Use following Lines of code for training the model on Cifar10 dataset.

# imports
import torch
from colourgan.config import get_cfg
from colourgan.data import Cifar10Dataset
from colourgan.model import ColourGAN

# Downloading the data and creating the dataloaders
dataloaders = Cifar10Dataset('cifar10', batch_size = 8).get_dataloaders()

# config
cfg = get_cfg()

# creating the model
model = ColourGAN(cfg)

# train
model.train(dataloaders['train'],dataloaders['test'], epochs = 100)

Pretrained Weights

Generator
Discriminator

References

GAN_image_colorizing

Colorizing-with-GANs

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

colourgan-1.0.3.tar.gz (10.8 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

colourgan-1.0.3-py3-none-any.whl (11.7 kB view details)

Uploaded Python 3

File details

Details for the file colourgan-1.0.3.tar.gz.

File metadata

  • Download URL: colourgan-1.0.3.tar.gz
  • Upload date:
  • Size: 10.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.1 importlib_metadata/4.5.0 pkginfo/1.7.0 requests/2.25.1 requests-toolbelt/0.9.1 tqdm/4.61.0 CPython/3.9.5

File hashes

Hashes for colourgan-1.0.3.tar.gz
Algorithm Hash digest
SHA256 87e2adc83c0becff0635bba393a2257e4868fa5f2e41487e4ead26b18df63138
MD5 2d1eb187d9b9bb2c065b1468467ee1ef
BLAKE2b-256 40dc45eda9f855eee08ae03379ed4b7a3718c86f592b439ecf8e93d19b139170

See more details on using hashes here.

File details

Details for the file colourgan-1.0.3-py3-none-any.whl.

File metadata

  • Download URL: colourgan-1.0.3-py3-none-any.whl
  • Upload date:
  • Size: 11.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.4.1 importlib_metadata/4.5.0 pkginfo/1.7.0 requests/2.25.1 requests-toolbelt/0.9.1 tqdm/4.61.0 CPython/3.9.5

File hashes

Hashes for colourgan-1.0.3-py3-none-any.whl
Algorithm Hash digest
SHA256 6ac07e671acd178f9fff1a0613187b04a934f8be71ee4918faa31a83d37e5993
MD5 5bedcee4283efafe4faf5abc6a7579ea
BLAKE2b-256 486c6c7358f7cdc42ee4112dafd5dac9b354ed3faf39abd99b48d381a1ceef41

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page