MBA-RainGAN: A Multi-Branch Attention Generative Adversarial Network for Mixture of Rain Removal

Shen, Yiyang; Feng, Yidan; Wang, Weiming; Dong, Liang; Qin, Jing; Xie, Haoran; Wei, Mingqiang

doi:10.1109/icassp43922.2022.9746588

Cited by 9 publications

(6 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They also used a depth-guided GAN to recover the background details. Additional approaches [15,16,36,37] have been proposed to address the mixture of rain. For instance, Wang et al [36] observed that rain streaks and rainy haze are intricately connected, while current rain image generation models fails to accurately model this property.…”

Section: Rain Streaks and Rainy Haze Removalmentioning

confidence: 99%

“…Hu et al [15,16] developed a rain imaging process based on the visual effects of rain in relation to scene depth and presented a depth-guided network to generate a rain-free image. Later, MBA-RainGAN [37], a multibranch attention generative adversarial network, was proposed to remove entangled rain streaks, rainy haze, and raindrops. However, these methods are only trained on synthetic rainy images and fall short in predicting an accurate depth map, thereby limiting their performance on real-world rainy images.…”

Section: Rain Streaks and Rainy Haze Removalmentioning

confidence: 99%

“…The rain streak removal methods comprise two traditional methods (DSC [4] and GMM [5]), with eight supervised learning methods (UMRL [49], SPA-Net [8], PreNet [28], DCSFN [50], MSPFN [11], MPRNet [12], DerainRLNet [51], and CCN [52]) and three semisupervised learning methods (SIRR [30], Syn2Real [13], and JRGR [14]). The rain streak and rainy haze removal methods comprise DAF-Net [16], DGNL-Net [15], and MBA-RainGAN [37]. Additionally, we compare our method with these rain streaks removal method combined with the FFA-Net dehazing method [53].…”

Section: Baselinesmentioning

confidence: 99%

See 2 more Smart Citations

Semi-RainGAN: A Semisupervised Coarse-to-Fine Guided Generative Adversarial Network for Mixture of Rain Removal

Yu,

Shu,

Zhang

et al. 2023

Symmetry

View full text Add to dashboard Cite

Images taken in various real-world scenarios meet the symmetrical goal of simultaneously removing foreground rain-induced occlusions and restoring the background details. This inspires us to remember the principle of symmetry; real-world rain is a mixture of rain streaks and rainy haze and degrades the visual quality of the background. Current efforts formulate image rain streak removal and rainy haze removal as separate models, which disrupts the symmetrical characteristics of real-world rain and background, leading to significant performance degradation. To achieve this symmetrical balance, we propose a novel semisupervised coarse-to-fine guided generative adversarial network (Semi-RainGAN) for the mixture of rain removal. Beyond existing wisdom, Semi-RainGAN is a joint learning paradigm of the mixture of rain removal and attention and depth estimation. Additionally, it introduces a coarse-to-fine guidance mechanism that effectively fuses estimated image, attention, and depth features. This mechanism enables us to achieve symmetrically high-quality rain removal while preserving fine-grained details. To bridge the gap between synthetic and real-world rain, Semi-RainGAN makes full use of unpaired real-world rainy and clean images, enhancing its generalization to real-world scenarios. Extensive experiments on both synthetic and real-world rain datasets demonstrate clear visual and numerical improvements of Semi-RainGAN over sixteen state-of-the-art models.

show abstract

Section: Rain Streaks and Rainy Haze Removalmentioning

confidence: 99%

Section: Rain Streaks and Rainy Haze Removalmentioning

confidence: 99%

Section: Baselinesmentioning

confidence: 99%

See 1 more Smart Citation

Semi-RainGAN: A Semisupervised Coarse-to-Fine Guided Generative Adversarial Network for Mixture of Rain Removal

Yu,

Shu,

Zhang

et al. 2023

Symmetry

View full text Add to dashboard Cite

show abstract

“…More recently, Shen et al. [25] also considered raindrops in the raining image model and proposed a multi‐branch attention GAN.…”

Section: Related Workmentioning

confidence: 99%

A two‐stage method for single image de‐raining based on attention smoothed dilated network

Fan

Zhao

et al. 2022

IET Image Processing

View full text Add to dashboard Cite

Rain can severely hamper the visibility of scene objects. Although existing deep learning methods have reported promising performance, they often fail to obtain satisfactory results in many practical situations, especially when the input image contains both rain streaks and haze-like degradation. In this paper, a new two-stage method based on attention smoothed dilated network (SDN) is proposed. Unlike most fully-supervised methods, the mixture of rain streaks and haze-like effects is considered in the model. The proposed method consists of two stages. First, a generative adversarial network guided by the rain-streak attention map is proposed to remove rain streaks, where a multi-stage attention module is used to accurately locate rain streaks in the generator. Second, haze-like effects are further removed through SDN with the same structure as the generator. Extensive experiments on multiple datasets show that the method outperforms the state-of-the-art in both objective evaluation and visual quality.

show abstract

“…The images captured in heavy rain always include some fog layers, and thus the adverse effect must be considered. In Cycle‐Derain, an unsupervised attention mechanism without alerting the background [4, 22] has been taken to address this issue.…”

Section: Cycle‐derain Modelmentioning

confidence: 99%

An unsupervised generative adversarial network for single image deraining

Song

Guo

et al. 2021

IET Image Processing

View full text Add to dashboard Cite

As the basis of image processing, single image deraining has always been a significant and challenging issue. Due to the lack of real rainy images and corresponding clean images, most deraining networks are trained by synthetic datasets, which makes the output images unsatisfactory in real applications. Besides, note that a heavy rainfall is typically accompanied with some fog. Although some deraining networks have been proposed to remove the rain streaks in the rainy images, the output images may still be blurred due to the accompanied fog. In this paper, these problems existing in single image deraining is comprehensively considered, and propose a Cycle-Derain network based on an unsupervised attention-guided mechanism. Specifically, the Cycle-Derain network takes advantage of generative adversarial networks with two mappings and the cycle consistency loss to train both unpaired rainy images and rain-free images. Moreover, it introduces an unsupervised attention-guided mechanism and exploits the loop-search positioning algorithm to deal with the details of rain and fog in images. Extensive experiments have been carried out, and the results show that the proposed Cycle-Derain network is preferable compared with other deraining networks, especially in term of rainy image restoration.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

MBA-RainGAN: A Multi-Branch Attention Generative Adversarial Network for Mixture of Rain Removal

Cited by 9 publications

References 15 publications

Semi-RainGAN: A Semisupervised Coarse-to-Fine Guided Generative Adversarial Network for Mixture of Rain Removal

Semi-RainGAN: A Semisupervised Coarse-to-Fine Guided Generative Adversarial Network for Mixture of Rain Removal

A two‐stage method for single image de‐raining based on attention smoothed dilated network

An unsupervised generative adversarial network for single image deraining

Contact Info

Product

Resources

About