Deep Image Prior: Difference between revisions
(Created page with "[https://dmitryulyanov.github.io/deep_image_prior Website]<br> [https://arxiv.org/abs/1711.10925 Paper]<br> [https://github.com/DmitryUlyanov/deep-image-prior Deep image prior...") |
No edit summary |
||
| Line 1: | Line 1: | ||
[https://dmitryulyanov.github.io/deep_image_prior Website]<br> | [https://dmitryulyanov.github.io/deep_image_prior Website]<br> | ||
[https://arxiv.org/abs/1711.10925 Paper]<br> | [https://arxiv.org/abs/1711.10925 Paper]<br> | ||
[https://github.com/DmitryUlyanov/deep-image-prior Deep image prior | [https://dmitryulyanov.github.io/assets/projects/pdf.svg Supplementary Material] | ||
[https://drive.google.com/open?id=15gUH7ebKmgsyhIWZFEIIsk4Q1uYc1GAF Mirror]<br> | |||
[https://github.com/DmitryUlyanov/deep-image-prior Deep image prior GitHub] | |||
The basic idea is that your choice of neural network imparts a prior on the output.<br> | The basic idea is that your choice of neural network imparts a prior on the output.<br> | ||
| Line 15: | Line 17: | ||
==Architectures== | ==Architectures== | ||
See the supplementary material | |||
Latest revision as of 16:33, 27 April 2020
Website
Paper
Supplementary Material
Mirror
Deep image prior GitHub
The basic idea is that your choice of neural network imparts a prior on the output.
That is, by using a CNN, the output of your neural network tends to be smoother images.
They show this by fitting a neural network to noise vs fitting it to natural or drawn images.
CNNs are harder to fit to noise than to natural images.
This leads to numerous applications including denoising, reconstruction, super-resolution, and impainting.
For each application, the idea is to train a CNN neural network to reconstruct that particular image.
However, you perform early stopping so that the neural network does not learn to reconstruct the noise or artifacts within the image.
Since CNNs find it easier to reconstruct low-frequency details which correspond too the image structure,
they'll reconstruct those first during the optimization process.
Architectures
See the supplementary material