Segmentation-Free Dynamic Scene Deblurring

By Tae Hyun Kim, Kyoung Mu Lee. 2014

192 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Hand-Held Video Deblurring Via Efficient Fourier Aggregation (2015, background): “, [10]– [12]) and we concentrate on general deblurring techniques that target (or can be easily adapted to) camera shake removal.”
Deep Multi-scale Convolutional Neural Network for Dynamic Scene Deblurring (2016, background): “In addition, Kim and Lee [15] approximated the blur kernel as locally linear and proposed an approach that estimates both the latent image and the local linear motions jointly.”
$L_0$ -Regularized Intensity and Gradient Prior for Deblurring Text Images and Beyond (2017, background): “In addition to camera motion, blurred images caused by different object motions are analyzed [28], [29], [30].”
Learning a convolutional neural network for non-uniform motion blur removal (2015, methodology): “Other approaches [13, 29] jointly estimate the sharp image and blur kernels using a sparsity prior.”
Real-World Blur Dataset for Learning and Benchmarking Deblurring Algorithms (2020, background): “To deal with more realistic blur, several approaches with extended blur models have been proposed, but their performance is still limited due to the inherent ill-posedness of the inverse problem [18, 44, 8, 14, 17, 19].”
From Motion Blur to Motion Flow: A Deep Learning Solution for Removing Heterogeneous Motion Blur (2016, methodology): “By using a linear motion model to indicate each pixel’s motion during imaging process [18], and letting p = (i, j) denote a pixel location, the motion at pixel p, can be represented by a 2dimensional motion vector (up, vp), where up and vp represent the movement in the horizontal and vertical direct”
Spatio-Temporal Filter Adaptive Network for Video Deblurring (2019, methodology): “Another group of methods [4, 37, 8, 9] use single or multiple images to estimate optical flow which is treated as the approximation of non-uniform blur kernels.”
Online Video Deblurring via Dynamic Temporal Blending Network (2017, methodology): “Kim and Lee [17] proposed a method to parametrize pixelwise varying kernels with motion flows in a single image, and they naturally extended it to deal with blurs in videos [18].”
Human-Aware Motion Deblurring (2019, methodology): “They deblur the background and object regions separately by relying on pre-generated segments [29, 16], or estimating motion flow to facilitate blur kernel prediction in a segmentation-free framework [17].”
Dynamic Scene Deblurring Using Spatially Variant Recurrent Neural Networks (2018, methodology): “Conventional methods [9, 10, 23] usually add constrains on the estimated image and blur kernel, and then optimize complex objective functions.”
Dynamic Scene Deblurring With Parameter Selective Sharing and Nested Skip Connections (2019, methodology): “Later, Kim and Lee [13] assumed that motion is locally linearly varying, and hence proposed a segmentation-free approach to handle this problem.”