Discriminative Transfer Learning for General Image Restoration.

Citation data:

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society, ISSN: 1941-0042, Vol: 27, Issue: 8, Page: 4091-4104

Publication Year:
2018
Usage 12
Downloads 8
Abstract Views 4
Captures 20
Readers 20
Social Media 12
Tweets 12
Repository URL:
http://hdl.handle.net/10754/626486; http://arxiv.org/abs/1703.09245
PMID:
29993740
DOI:
10.1109/tip.2018.2831925
Author(s):
Xiao, Lei; Heide, Felix; Heidrich, Wolfgang; Schölkopf, Bernhard; Hirsch, Michael
Publisher(s):
Institute of Electrical and Electronics Engineers (IEEE)
Tags:
Computer Science; discriminative learning; Image restoration; proximal optimization; Computer Science - Computer Vision and Pattern Recognition
Most Recent Tweet View All Tweets
article description
Recently, several discriminative learning approaches have been proposed for effective image restoration, achieving convincing trade-off between image quality and computational efficiency. However, these methods require separate training for each restoration task (e.g., denoising, deblurring, demosaicing) and problem condition (e.g., noise level of input images). This makes it time-consuming and difficult to encompass all tasks and conditions during training. In this paper, we propose a discriminative transfer learning method that incorporates formal proximal optimization and discriminative learning for general image restoration. The method requires a single-pass discriminative training and allows for reuse across various problems and conditions while achieving an efficiency comparable to previous discriminative approaches. Furthermore, after being trained, our model can be easily transferred to new likelihood terms to solve untrained tasks, or be combined with existing priors to further improve image restoration quality.