PROJECT TITLE :

Discriminative Transfer Learning for General Image Restoration - 2018

ABSTRACT:

Recently, several discriminative learning approaches are proposed for effective image restoration, achieving convincing tradeoff between image quality and computational efficiency. However, these ways require separate training for every restoration task (e.g., denoising, deblurring, and demosaicing) and downside condition (e.g., noise level of input images). This makes it time-consuming and difficult to encompass all tasks and conditions during coaching. During this Project, we have a tendency to propose a discriminative transfer learning technique that includes formal proximal optimization and discriminative learning for general image restoration. The tactic needs a single-pass discriminative training and allows for reuse across varied issues and conditions whereas achieving an efficiency comparable to previous discriminative approaches. Furthermore, after being trained, our model can be simply transferred to new chance terms to solve untrained tasks, or be combined with existing priors to more improve image restoration quality.