On-Demand Learning for Deep Image Restoration

12/05/2016
by   Ruohan Gao, et al.
0

While machine learning approaches to image restoration offer great promise, current methods risk training models fixated on performing well only for image corruption of a particular level of difficulty---such as a certain level of noise or blur. First, we examine the weakness of conventional "fixated" models and demonstrate that training general models to handle arbitrary levels of corruption is indeed non-trivial. Then, we propose an on-demand learning algorithm for training image restoration models with deep convolutional neural networks. The main idea is to exploit a feedback mechanism to self-generate training instances where they are needed most, thereby learning models that can generalize across difficulty levels. On four restoration tasks---image inpainting, pixel interpolation, image deblurring, and image denoising---and three diverse datasets, our approach consistently outperforms both the status quo training procedure and curriculum learning alternatives.

READ FULL TEXT

page 1

page 7

page 12

page 15

page 16

page 17

page 18

page 19

research
04/17/2019

Modulating Image Restoration with Continual Levels via Adaptive Feature Modification Layers

In image restoration tasks, learning from discrete and fixed restoration...
research
11/10/2022

ClassPruning: Speed Up Image Restoration Networks by Dynamic N:M Pruning

Image restoration tasks have achieved tremendous performance improvement...
research
04/23/2019

Path-Restore: Learning Network Path Selection for Image Restoration

Very deep Convolutional Neural Networks (CNNs) have greatly improved the...
research
03/27/2017

Discriminative Transfer Learning for General Image Restoration

Recently, several discriminative learning approaches have been proposed ...
research
01/18/2022

Dilated Convolutional Neural Networks for Lightweight Diacritics Restoration

Diacritics restoration has become a ubiquitous task in the Latin-alphabe...
research
09/07/2020

Are Deep Neural Architectures Losing Information? Invertibility Is Indispensable

Ever since the advent of AlexNet, designing novel deep neural architectu...
research
01/13/2014

Insights into analysis operator learning: From patch-based sparse models to higher-order MRFs

This paper addresses a new learning algorithm for the recently introduce...

Please sign up or login with your details

Forgot password? Click here to reset