arXiv Analytics

Sign in

arXiv:1712.02327 [cs.CV]AbstractReferencesReviewsResources

Burst Denoising with Kernel Prediction Networks

Ben Mildenhall, Jonathan T. Barron, Jiawen Chen, Dillon Sharlet, Ren Ng, Robert Carroll

Published 2017-12-06Version 1

We present a technique for jointly denoising bursts of images taken from a handheld camera. In particular, we propose a convolutional neural network architecture for predicting spatially varying kernels that can both align and denoise frames, a synthetic data generation approach based on a realistic noise formation model, and an optimization guided by an annealed loss function to avoid undesirable local minima. Our model matches or outperforms the state-of-the-art across a wide range of noise levels on both real and synthetic data.

Related articles: Most relevant | Search more
arXiv:1703.05593 [cs.CV] (Published 2017-03-16)
Convolutional neural network architecture for geometric matching
arXiv:1511.00561 [cs.CV] (Published 2015-11-02)
SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation
arXiv:1804.00497 [cs.CV] (Published 2018-03-28, updated 2018-10-03)
MicronNet: A Highly Compact Deep Convolutional Neural Network Architecture for Real-time Embedded Traffic Sign Classification