Semantic segmentation with greyscale images

Question

I'm trying to reproduce a research with greyscale images instead of colour images.

I have found that there are pre-trained networks, like VGG16, with ImageNet. But that dataset has colour images, and I can't use it because I'm going to use greyscale images.

Is there any pre-trained network with greyscale images?

Failing that, I can also train a network with a greyscale image dataset but I can't find any.

Your problem has been asked and resolved in this link: https://stackoverflow.com/questions/51995977/how-can-i-use-a-pre-trained-neural-network-with-grayscale-images — , Jun 25 '21 at 12:07

score 0 · Answer 1 · answered Sep 15 '20 at 20:26

0

Use ImageNet, convert every image to grayscale and train VGG16.

answered Sep 15 '20 at 20:26

score 0 · Answer 2 · answered May 22 '21 at 08:37

0

Your problem is that your dataset has one value per pixel, whereas ImageNet expects 3? Just convert your data to "color images" by passing the same value on all 3 (RGB) channels.

answered May 22 '21 at 08:37

Paul

903
4
10

Semantic segmentation with greyscale images

2 Answers2