Translution-SNet: A Semisupervised Hyperspectral Image Stripe Noise Removal Based on Transformer and CNN

Wang, Chengjun; Xu, Miaozhong; Jiang, Yonghua; Zhang, Guo; Cui, Hao; Li, Litao; Li, Da

doi:10.1109/tgrs.2022.3182745

Cited by 22 publications

(9 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, HSI-specific networks have also been developed to take advantage of both the spectral and spatial properties of HSI [11,10,37,38,39,40,41,42,43]. In particular, SS-CAN [11] is an HSI-specific denoising network, that combines group convolutions and attention modules to effectively exploit spatial and spectral information in images.…”

Section: Deep Learning Approaches To Hsi Denoisingmentioning

confidence: 99%

See 1 more Smart Citation

Hyperspectral image denoising via self-modulating convolutional neural networks

Torun,

Yuksel,

Erdem

et al. 2024

Signal Processing

View full text Add to dashboard Cite

Section: Deep Learning Approaches To Hsi Denoisingmentioning

confidence: 99%

“…More recently, attention-based methods were proposed to capture non-local features [39,40,41,42]. Yuan et.…”

Section: Deep Learning Approaches To Hsi Denoisingmentioning

confidence: 99%

Hyperspectral image denoising via self-modulating convolutional neural networks

Torun,

Yuksel,

Erdem

et al. 2024

Signal Processing

View full text Add to dashboard Cite

“…2(a), stripe noise primarily disrupts the image's horizontal gradient, which leads to stripe noise accumulating in the low and horizontal high-frequency components of the HDWT. Consequently, the HDWT has gained widespread adoption as a natural stripe feature extractor for stripe removal [36]. For example, Guan et al [26] first combined DWT and CNN for the destriping task.…”

Section: B Deep Wavelet Transform-based Image Processingmentioning

confidence: 99%

Wavelet Channel Attention Module With A Fusion Network For Single Image Deraining

Yang¹,

Yang²,

Wang³

2020

2020 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

Single image deraining is a crucial problem because rain severely degenerates the visibility of images and affects the performance of computer vision tasks like outdoor surveillance systems and intelligent vehicles. In this paper, we propose the new convolutional neural network (CNN) called the wavelet channel attention module with a fusion network. Wavelet transform and the inverse wavelet transform are substituted for down-sampling and up-sampling so feature maps from the wavelet transform and convolutions contain different frequencies and scales. Furthermore, feature maps are integrated by channel attention. Our proposed network learns confidence maps of four sub-band images derived from the wavelet transform of the original images. Finally, the clear image can be well restored via the wavelet reconstruction and fusion of the low-frequency part and high-frequency parts. Several experimental results on synthetic and real images present that the proposed algorithm outperforms state-of-theart methods.

show abstract

“…Islam et al [19] used a CNN to build a four-stage model and applied a migration learning mechanism for faster training. Wang et al [20] proposed Translution-SNet, which is based on a semi-supervised training strategy using convolution and transformer approaches for feature extraction to improve the CNN's ability to handle various complex noise types. Lyu et al [21] presented an adversarial network for mixed noise removal; this method uses a competing generator network and a discriminator network in combination with a feature extractor network for additional training.…”

Section: Introductionmentioning

confidence: 99%

“…Wang et al. [20] proposed Translution‐SNet, which is based on a semi‐supervised training strategy using convolution and transformer approaches for feature extraction to improve the CNN's ability to handle various complex noise types. Lyu et al.…”

Section: Introductionmentioning

confidence: 99%

SDAUNet: A simple dual attention mechanism UNet for mixed noise removal

Jiang

Hong

Zhao

et al. 2023

IET Image Processing

View full text Add to dashboard Cite

Convolutional neural networks (CNNs) have demonstrated impressive results in additive white Gaussian noise removal due to their strong fitting ability. However, their performance in mixed noise removal remains unsatisfactory, primarily due to their limited receptive field that focuses only on the local features of images and disregards global information. To ameliorate this issue, recent state‐of‐the‐art approaches employ attention mechanism (AM) to capture the global information. However, most AM based methods still suffer from low computational efficiency. In this paper, a novel model named simple dual attention mechanism UNet (SDAUNet) for mixed noise removal is proposed. In SDAUNet, the UNet architecture is used to gradually acquire multi‐scale image features and provide a more comprehensive and accurate representation of the image features than other CNNs. A simple dual attention convolutional block is presented to acquire the global image features that can successfully capture image details with a low burden. The experimental results demonstrate that the SDAUNet model can achieve better measurement metrics and visual performance than other state‐of‐the‐art methods.

show abstract

Translution-SNet: A Semisupervised Hyperspectral Image Stripe Noise Removal Based on Transformer and CNN

Cited by 22 publications

References 57 publications

Hyperspectral image denoising via self-modulating convolutional neural networks

Hyperspectral image denoising via self-modulating convolutional neural networks

Wavelet Channel Attention Module With A Fusion Network For Single Image Deraining

SDAUNet: A simple dual attention mechanism UNet for mixed noise removal

Contact Info

Product

Resources

About