← all work generative-vision-lab

Colourisation

Predict colour (ab) from luminance (L) in Lab space.

Result

splitPSNR
train n=42822.5389
val n=42822.2671
test n=14221.9107

The finding

Notes

- Trained from scratch for 100 epochs on 2,280 images, predicting the ab channels from L in Lab space. - The dataset was filtered for colour content: near-grey and low-saturation images were dropped when it was built.

Figures

00729_after_output.png
00729_after_output
00729_before_output.png
00729_before_output
00729_ground_truth.png
00729_ground_truth
00729_input_grey.png
00729_input_grey
00933_after_output.png
00933_after_output
00933_before_output.png
00933_before_output
00933_ground_truth.png
00933_ground_truth
00933_input_grey.png
00933_input_grey
01231_after_output.png
01231_after_output
01231_before_output.png
01231_before_output
01231_ground_truth.png
01231_ground_truth
01231_input_grey.png
01231_input_grey

Elsewhere

Code and full results on GitHub

Every number is read from the file the run produced. github.com/hammasbuilds · kaggle.com/muhammadhammas13