A Neural Approach to Blind Motion Deblurring

Abstract

We present a new method for blind motion deblurring that uses a neural network trained to compute estimates of sharp image patches from observations that are blurred by an unknown motion kernel. Instead of regressing directly to patch intensities, this network learns to predict the complex Fourier coefficients of a deconvolution filter to be applied to the input patch for restoration. For inference, we apply the network independently to all overlapping patches in the observed image, and average its outputs to form an initial estimate of the sharp image. We then explicitly estimate a single global blur kernel by relating this estimate to the observed image, and finally perform non-blind deconvolution with this kernel. Our method exhibits accuracy and robustness close to state-of-the-art iterative methods, while being much faster when parallelized on GPU hardware.

ECCV 2016. Project page at http://www.ttic.edu/chakrabarti/ndeblur

References in corpus (1)

• Learning a Convolutional Neural Network for Non-uniform Motion Blur Removal

Cited by in corpus (18)

• Examining the Impact of Blur on Recognition by Convolutional Networks
• Deep Video Deblurring
• Deep Multi-scale Convolutional Neural Network for Dynamic Scene Deblurring
• Online Video Deblurring via Dynamic Temporal Blending Network
• Scale-recurrent Network for Deep Image Deblurring
• Learning a Discriminative Prior for Blind Image Deblurring
• Blind Deblurring using Deep Learning: A Survey
• Learning to Extract a Video Sequence from a Single Motion-Blurred Image
• Deep Semantic Face Deblurring
• Learning a Dilated Residual Network for SAR Image Despeckling
• DeepDeblur: Fast one-step blurry face images restoration
• End-to-End Learning for Image Burst Deblurring
• InverseNet: Solving Inverse Problems with Splitting Networks
• Reblur2Deblur: Deblurring Videos via Self-Supervised Learning
• Learning Fully Convolutional Networks for Iterative Non-blind Deconvolution
• Deep joint rain and haze removal from single images
• Robust Blind Deconvolution via Mirror Descent
• Iterative Residual Image Deconvolution