eduzhai > Applied Sciences > Engineering >

Flexible Image Denoising with Multi-layer Conditional Feature Modulation

  • Save

... pages left unread,continue reading

Document pages: 12 pages

Abstract: For flexible non-blind image denoising, existing deep networks usually takeboth noisy image and noise level map as the input to handle various noiselevels with a single model. However, in this kind of solution, the noisevariance (i.e., noise level) is only deployed to modulate the first layer ofconvolution feature with channel-wise shifting, which is limited in balancingnoise removal and detail preservation. In this paper, we present a novelflexible image enoising network (CFMNet) by equipping an U-Net backbone withmulti-layer conditional feature modulation (CFM) modules. In comparison tochannel-wise shifting only in the first layer, CFMNet can make better use ofnoise level information by deploying multiple layers of CFM. Moreover, each CFMmodule takes onvolutional features from both noisy image and noise level map asinput for better trade-off between noise removal and detail preservation.Experimental results show that our CFMNet is effective in exploiting noiselevel information for flexible non-blind denoising, and performs favorablyagainst the existing deep image denoising methods in terms of both quantitativemetrics and visual quality.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×