SDWNet: A Straight Dilated Network with Wavelet Transformation for image Deblurring

Zou, Wenbin; Jiang, Mingchao; Zhang, Yuncheng; Chen, Liang; Lu, Zhiyong; Wu, Yi

doi:10.1109/iccvw54120.2021.00216

Cited by 70 publications

(34 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, the characteristics of image features at different scales are neglected in them. Besides, for the methods which considered the frequency information of images in deblurring task [31,62], they either only focused on high-frequency features or treated different frequency features of an image with the same strategy. Hence, they are still different from our proposed method.…”

Section: Comparison With Other Methodsmentioning

confidence: 99%

“…In image-to-image regression framework, Liu et al [31] designed a two-stage method which first separates the high-frequency residual information from the blurry image and then adopt an encoderdecoder network to realize the high frequency information refinement. Zou et al [62] utilized discrete wavelet transform to divide the dilated convolution features into four frequency bands, so that different frequency features can be refined independently. Nevertheless, the above two methods only separate the image frequency in the first or last layer of the network.…”

Section: Frequency Separationmentioning

confidence: 99%

“…But it only focuses the high-frequency image features. Although SDWNet [62] utilizes the wavelet transformation for image frequency separation, it processes the low-and highfrequency image features using the same network. Hence, the deblurring results obtained by the above two methods are still not optimal.…”

Section: Quantitative and Qualitative Evaluation 421 Gopro Datasetmentioning

confidence: 99%

“…Hence, the deblurring results obtained by the above two methods are still not optimal. Besides, the methods in [62] and [31] only capture image frequency information at the first or last layer of the network and ignore the multi-scale frequency features of the image, which may also decrease their performance. At last, the proposed MSFS-Net can capture different frequency features of the image at multiple scales and designs different strategies (i.e., contrast learning and consistent loss) to handle them separately.…”

Section: Quantitative and Qualitative Evaluation 421 Gopro Datasetmentioning

confidence: 99%

“…Therefore, the above methods may suffer from both high computational burden and inaccurate blur kernel estimation. More recently, some other CNN based image deblurring methods were proposed to directly learn the relationship between blurry and sharp images by an image-to-image regression manner [1,2,4,8,29,31,32,49,62]. Compared with other works, the advantage of image-to-image regression methods is that they could avoid the deblurring errors induced by inadequate blur kernel estimation.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Multi-scale frequency separation network for image deblurring

Zhang¹,

Li²,

Miao³

et al. 2022

Preprint

View full text Add to dashboard Cite

Image deblurring aims to restore the detailed texture information or structures from the blurry images, which has become an indispensable step in many computer-vision tasks. Although various methods have been proposed to deal with the image deblurring problem, most of them treated the blurry image as a whole and neglected the characteristics of different image frequencies. In this paper, we present a new method called multi-scale frequency separation network (MSFS-Net) for image deblurring. MSFS-Net introduces the frequency separation module (FSM) into an encoder-decoder network architecture to capture the lowand high-frequency information of image at multiple scales. Then, a simple cycle-consistency strategy and a sophisticated contrastive learning module (CLM) are respectively designed to retain the low-frequency information and recover the high-frequency information during deblurring. At last, the features of different scales are fused by a crossscale feature fusion module (CSFFM). Extensive experiments on benchmark datasets show that the proposed network achieves state-of-the-art performance.

show abstract

Section: Comparison With Other Methodsmentioning

confidence: 99%

Section: Frequency Separationmentioning

confidence: 99%

Section: Quantitative and Qualitative Evaluation 421 Gopro Datasetmentioning

confidence: 99%

Section: Quantitative and Qualitative Evaluation 421 Gopro Datasetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Multi-scale frequency separation network for image deblurring

Zhang¹,

Li²,

Miao³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

Frequency and Spatial Dual Guidance for Image Dehazing

Zheng

Zhou

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Image dehazing is quite challenging in dense-haze scenarios, where quite less original information remains in the hazy image. Though previous methods have made marvelous progress, they still suffer from information loss in content and color in dense-haze scenarios. The recently emerged Denoising Diffusion Probabilistic Model (DDPM) exhibits strong generation ability, showing potential for solving this problem. However, DDPM fails to consider the physics property of dehazing task, limiting its information completion capacity. In this work, we propose DehazeDDPM: A DDPM-based and physics-aware image dehazing framework that applies to complex hazy scenarios. Specifically, DehazeDDPM works in two stages. The former stage physically models the dehazing task with the Atmospheric Scattering Model (ASM), pulling the distribution closer to the clear data and endowing DehazeDDPM with fog-aware ability. The latter stage exploits the strong generation ability of DDPM to compensate for the haze-induced huge information loss, by working in conjunction with the physical modelling. Extensive experiments demonstrate that our method attains state-of-theart performance on both synthetic and real-world hazy datasets.

show abstract