FormNet : Formatted Learning for Image Restoration
In this paper, we propose a deep CNN to tackle the image restoration problem by learning formatted information. Previous deep learning based methods directly learn the mapping from corrupted images to clean images, and may suffer from the gradient exploding/vanishing problems of deep neural networks...
Ausführliche Beschreibung
Bibliographische Detailangaben
| Veröffentlicht in: | IEEE transactions on image processing : a publication of the IEEE Signal Processing Society. - 1992. - (2020) vom: 01. Mai
|
|---|
| 1. Verfasser: |
Jiao, Jianbo
(VerfasserIn) |
|---|
| Weitere Verfasser: |
Tu, Wei-Chih,
Liu, Ding,
He, Shengfeng,
Lau, Rynson W H,
Huang, Thomas S |
|---|
| Format: | Online-Aufsatz
|
|---|
| Sprache: | English |
|---|
| Veröffentlicht: |
2020
|
|---|
| Zugriff auf das übergeordnete Werk: | IEEE transactions on image processing : a publication of the IEEE Signal Processing Society
|
|---|
| Schlagworte: | Journal Article |
|---|