How to output an image with a CNN?

Question

I'm trying to do depth estimation with CNNs (this is my ultimate goal), but a problem that i found is: I just did image classifications with CNNs, using for example "CIFAR-10", "MNIST", "Cats vs Dogs", etc. To do depth estimation I need to output a new image (the NYUv2 dataset has the labeled images). So, I'll input an image like 256x256x3 and need to output another image with for example 228x228x3.

What I need to do? Can I just do the convolutions for a while and after that decrease the features maps and increase the dimension? Thanks

obs: I'm using Tensorflow 2.0

It depends on what you need to do but there are several ways to do image-to-image neural networks. You have for example U-nets (arxiv.org/abs/1505.04597), Res-nets (arxiv.org/abs/1608.03981), Auto-encoders, and much more. For single-image depth resolution, some people used very complex networks: cv-foundation.org/openaccess/content_cvpr_2015/papers/… — Zaccharie Ramzi
– Zaccharie Ramzi, Commented Dec 22, 2019 at 11:13

bsquare · Accepted Answer · 2020-04-24 06:55:13Z

0

I suggest you use a type of UNet. This kind of architecture has downsampling layers, followed by up sampling layers to get back to the original spatial dimensions.

edited Apr 24, 2020 at 6:55

answered Apr 22, 2020 at 10:06

bsquare

9966 silver badges10 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

How to output an image with a CNN?

1 Answer 1

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related