Unfortunately submitting just “2” to SOUSEN produces decent results, as everyone knows at this point. So I did that. Also submitting baseline to OBJDE gave a decent ranking 
For IMGCOL, DeOldify produces very good results. Link:
The script I used to run DeOldify:
from deoldify import device
from deoldify.device_id import DeviceId
#choices: CPU, GPU0…GPU7
device.set(device=DeviceId.GPU0)
from deoldify.visualize import *
plt.style.use(‘dark_background’)
torch.backends.cudnn.benchmark=True
import warnings
warnings.filterwarnings(“ignore”, category=UserWarning, message=".?Your .? set is empty.*?")
colorizer = get_image_colorizer(artistic=True)
render_factor=45
m=os.listdir(’./test_back_white_images’)
for file in m:
source_path = f"./test_back_white_images/{file}"
result_path = f"./test_color_images/{file}"
source_url=None
if source_url is not None:
result_path = colorizer.plot_transformed_image_from_url(url=source_url, path=source_path, render_factor=render_factor, compare=True)
else:
result_path = colorizer.plot_transformed_image(path=source_path, render_factor=render_factor, compare=True)
plt.close(‘all’)