English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Image demosaicking and denoising are the two key fundamental steps in digitalcamera pipelines, aiming to reconstruct clean color images from noisy luminancereadings. In this paper, we propose and study Wild-JDD, a novel learningframework for joint demosaicking and denoising in the wild. In contrast toprevious works which generally assume the ground truth of training data is aperfect reflection of the reality, we consider here the more common imperfectcase of ground truth uncertainty in the wild. We first illustrate itsmanifestation as various kinds of artifacts including zipper effect, colormoire and residual noise. Then we formulate a two-stage data degradationprocess to capture such ground truth uncertainty, where a conjugate priordistribution is imposed upon a base distribution. After that, we derive anevidence lower bound (ELBO) loss to train a neural network that approximatesthe parameters of the conjugate prior distribution conditioned on the degradedinput. Finally, to further enhance the performance for out-of-distributioninput, we design a simple but effective fine-tuning strategy by taking theinput as a weakly informative prior. Taking into account ground truthuncertainty, Wild-JDD enjoys good interpretability during optimization.Extensive experiments validate that it outperforms state-of-the-art schemes onjoint demosaicking and denoising tasks on both synthetic and realistic rawdatasets.

Joint Demosaicking and Denoising in the Wild: The Case of Training Under Ground Truth Uncertainty