Recurrent Conditional Generative Adversarial Network for Image Deblurring

Liu, Jing; Sun, Wanning; Li, Mengjie

doi:10.1109/access.2018.2888885

Cited by 24 publications

(9 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nah et al [15] used pixel wise MSE loss to train the network, and further combine MSE loss with adversarial loss of GAN to encourage the network to overcome over-smoothing. Liu et al [21] proposed the loss function of discriminator at three levels to evaluate recovered images with both global and local scopes. Similar methods based on GAN to address the problem are also reported [18]- [20].…”

Section: Related Workmentioning

confidence: 99%

“…Kupyn et al [20] augmented perceptual loss with Wasserstein adversarial loss to train a GAN. Liu et al [21] proposed an adversarial loss function based on Wasserstein GAN [37], which includes MSE loss for the generator to generate sharp structures and gradient penalty for the discriminator to constrain weights. Compared to these improved MSE losses, our loss function of the generator can preserve edge details and facilitate sharp edge prediction.…”

Section: B Analysis Of the Proposed Network 1) Ablation Study Of Thementioning

confidence: 99%

“…It is of great importance to capture long-range dependency in image deblurring based on deep learning. Many CNN-based deblurring methods develop various networks formed by large receptive fields to capture long-range dependencies, such as developing a deep multiscale network [15]- [17] or a generative adversarial network (GAN) [18]- [21]. These methods are used to cover large receptive fields, and to create solutions that are close to natural motion problems.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Deep Pyramid Generative Adversarial Network With Local and Nonlocal Similarity Features for Natural Motion Image Deblurring

Zhao

Gong

2019

IEEE Access

View full text Add to dashboard Cite

It is of great importance to capture long-range dependency in image deblurring based on deep learning. Existing methods often capture long-range dependency by a large receptive field, which contributes by deep stacks of local convolutional operations. Therefore, it restricts network representation ability and causes unpleasant artifacts of restored images. In this paper, we propose a deep pyramid generative adversarial network with local and nonlocal similarity features, called LNL-PGAN, for natural motion image deblurring. First, we propose a nonlocal feature block as an essential component of the pyramid generator for obtaining nonlocal similarity features at multiple levels. Second, we design a local feature block as another essential component to make a great balance between local and nonlocal similarity features. The local and nonlocal feature blocks capture meaningful short-range and long-range dependencies in the pyramid generator to increase network representation ability. Third, we design a multiscale generative adversarial loss to preserve edge details and facilitate sharp edge prediction of restored images, and we introduce a multistage training strategy to facilitate network training, which can further improve the quality of the restored image. Extensive experimental results demonstrate that the proposed method yields superior performance against state-of-the-art methods on natural motion image deblurring in terms of visual quality and objective index, and it can be used as a unified network for single and dynamic motion image deblurring. INDEX TERMS Motion deblurring, long-range dependency, nonlocal similarity feature, generative adversarial network. WEIHONG LI received the Ph.D. degree from Chongqing University, in 2006. She is currently a Professor with Chongqing University, China. Her current research interests are in the areas of pattern recognition and image processing.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: B Analysis Of the Proposed Network 1) Ablation Study Of Thementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Pyramid Generative Adversarial Network With Local and Nonlocal Similarity Features for Natural Motion Image Deblurring

Zhao

Gong

2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…However, in the real world, the blur may also be caused by object motion, camera shake and depth change, resulting in different pixels obtaining different motion trajectory. The other is the learnbased method, from the initial method of estimating the blur kernel by learning and then deconvoluting [5,6,7,8], to the end-to-end method to directly estimate the clear image from the fuzzy image [10,11,12,13,39,40]. It is proved that this method which transforms the deblurring problem into the image translation problem gets better results.…”

Section: Introductionmentioning

confidence: 99%

A Deep Motion Deblurring Network Using Channel Adaptive Residual Module

et al. 2021

View full text Add to dashboard Cite

In this paper, we solve the problem of dynamic scenes deblurring with motion blur. Restoration of images in the presence of motion blur necessitates a network design that the receptive field can completely cover all areas that need to be deblurred, while the existing network increases the receptive field by continuously stacking the ordinary convolutional layer or increasing the size of the convolution kernel. However, these methods inevitably increase the computational burden of the network. We propose a novel architecture consisting of a channel adaptive residual module. Different features of the blurred image are extracted and distributed on each feature channel. Our network can calculate the weight of each channel through learning, and extract the image features adaptively according to different degrees of blurring and importance of information. We embed the module in a modified encoder-decoder design with skip connections to achieve multi-scale feature fusion for further performance improvement. The extensive comparison with the existing techniques in the baseline dynamic scene deblurring dataset shows that the proposed network can effectively realize image deblurring, and the accuracy and speed are comparable with the existing techniques.

show abstract

“…They integrate encoder and decoder networks into a single framework, and adopt end-to-end training strategy to make deblurring task simpler. Other deep neural structures, such as recurrent neural network (RNN) or generative adversarial network (GAN), were applied to deal with deblurring problem [29]- [31].…”

Section: Introductionmentioning

confidence: 99%

A Motion Deblur Method Based on Multi-Scale High Frequency Residual Image Learning

et al. 2020

View full text Add to dashboard Cite

Non-uniform blind deblurring of dynamic scenes has always been a challenging problem in image processing because of the diverse of blurring sources. Traditional methods based on energy minimization cannot make accurate kernel estimation. It leads to that some high frequency details cannot be fully recovered. Recently, many methods based on convolution neural networks (CNNs) have been proposed to improve the overall performance. Followed by this trend, we first propose a two-stage deblurring module to recover the blur images of dynamic scenes based on high frequency residual image learning. The first stage performs initial deburring with the blur kernel estimated by the salient structure. The second stage calculates the difference of input image and initially deblurred image, referred to as residual image, and adopt an encoder-decoder network to refine the residual image. Finally, we can combine the refined residual image with the input blurred image to obtain the latent image. To increase deblurring performance, we further propose a coarse-to-fine framework based on the deblurring module. It performs the deblurring module many times in a multi-scale manner which can gradually restore the sharp edge details of different scales. Experiments conducted on three benchmark datasets demonstrate the proposed method achieves competitive performance of state-of-art methods.INDEX TERMS Image deblurring, dynamic blur, non-uniform blind deblurring, deep learning.

show abstract

Recurrent Conditional Generative Adversarial Network for Image Deblurring

Cited by 24 publications

References 31 publications

Deep Pyramid Generative Adversarial Network With Local and Nonlocal Similarity Features for Natural Motion Image Deblurring

Deep Pyramid Generative Adversarial Network With Local and Nonlocal Similarity Features for Natural Motion Image Deblurring

A Deep Motion Deblurring Network Using Channel Adaptive Residual Module

A Motion Deblur Method Based on Multi-Scale High Frequency Residual Image Learning

Contact Info

Product

Resources

About