Unsupervised Denoising for Satellite Imagery using Wavelet Subband CycleGAN

Song, Joonyoung; Jeong, Jae-Heon; Park, Dae-Soon; Kim, Hyun-Ho; Seo, Doo-Chun; Ye, Jong Chul

doi:10.48550/arxiv.2002.09847

Cited by 3 publications

(6 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This process can be easily conducted by zeroing the last LL band. Our method is different from the previous method [22] in which only horizontal or vertical bands among wavelet subbands are utilized for network training.…”

Section: Wavelet High-frequency Image For Low-dose Ct Denoisingmentioning

confidence: 94%

“…Yet another advantage of using a wavelet transform for deep learning is that the directional components to be modified can be separated from the other bands based on the prior knowledge, so that the network can easily focus on learning the noise components. For example, in a recent paper by Song et al [22], only a subset of the wavelet bands was used to recompose a wavelet directional image, and the neural network was designed to learn the mapping between these directional images in an unsupervised manner using cycleGAN.…”

Section: B Deep Learning Using Wavelet High Frequency Imagesmentioning

confidence: 99%

“…3. The method is similar to the method introduced in the study by Song et al [22]. First, the low-dose CT image is decomposed repeatedly using wavelet transform.…”

Section: Wavelet High-frequency Image For Low-dose Ct Denoisingmentioning

confidence: 99%

“…To further improve the performance, our method is com-bined with the wavelet residual proccessing. More specifically, Song et al [22] have shown that the preprocessing method using wavelet transform is effective in separating the common components of an image from noises, so that the CNN can learn the noise better. Accordingly, the network trained using wavelet high-frequency detail images can improve noise reduction performance and preserve the information in the original CT image.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

AdaIN-Based Tunable CycleGAN for Efficient Unsupervised Low-Dose CT Denoising

2021

IEEE Trans. Comput. Imaging

Self Cite

View full text Add to dashboard Cite

Recently, deep learning approaches have been extensively studied for low-dose CT denoising thanks to its superior performance despite the fast computational time. In particular, cycleGAN has been demonstrated as a powerful unsupervised learning scheme to improve the low-dose CT image quality without requiring matched high-dose reference data. Unfortunately, one of the main limitations of the cycleGAN approach is that it requires two deep neural network generators at the training phase, although only one of them is used at the inference phase. The secondary auxiliary generator is needed to enforce the cycle-consistency, but the additional memory requirement and increases of the learnable parameters are the main huddles for cycleGAN training. To address this issue, here we propose a novel cycleGAN architecture using a single switchable generator. In particular, a single generator is implemented using adaptive instance normalization (AdaIN) layers so that the baseline generator converting a low-dose CT image to a routine-dose CT image can be switched to a generator converting high-dose to lowdose by simply changing the AdaIN code. Thanks to the shared baseline network, the additional memory requirement and weight increases are minimized, and the training can be done more stably even with small training data. Experimental results show that the proposed method outperforms the previous cycleGAN approaches while using only about half the parameters.

show abstract

Section: Wavelet High-frequency Image For Low-dose Ct Denoisingmentioning

confidence: 94%

Section: B Deep Learning Using Wavelet High Frequency Imagesmentioning

confidence: 99%

“…3. The method is similar to the method introduced in the study by Song et al [22]. First, the low-dose CT image is decomposed repeatedly using wavelet transform.…”

Section: Wavelet High-frequency Image For Low-dose Ct Denoisingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

AdaIN-Based Tunable CycleGAN for Efficient Unsupervised Low-Dose CT Denoising

2021

IEEE Trans. Comput. Imaging

Self Cite

View full text Add to dashboard Cite

show abstract

“…CycleGAN has shown great performance especially in unsupervised image artifact removal. Kang et al [32] proposed a CycleGAN-based-model for the removal of Low-Dose CT noise, and Song et al [33] also proposed a CycleGAN-basedmodel for the removal of noise in satellite imagery. Given the success, one is interested whether the resulting improvement is real or cosmetic changes.…”

Section: A Geometry Of Cycleganmentioning

confidence: 99%

Unsupervised CT Metal Artifact Learning using Attention-guided beta-CycleGAN

Lee¹,

Gu²,

Ye³

2020

Preprint

Self Cite

View full text Add to dashboard Cite

Metal artifact reduction (MAR) is one of the most important research topics in computed tomography (CT). With the advance of deep learning technology for image reconstruction, various deep learning methods have been also suggested for metal artifact removal, among which supervised learning methods are most popular. However, matched non-metal and metal image pairs are difficult to obtain in real CT acquisition. Recently, a promising unsupervised learning for MAR was proposed using feature disentanglement, but the resulting network architecture is complication and difficult to handle large size clinical images. To address this, here we propose a much simpler and much effective unsupervised MAR method for CT. The proposed method is based on a novel β-cycleGAN architecture derived from the optimal transport theory for appropriate feature space disentanglement. Another important contribution is to show that attention mechanism is the key element to effectively remove the metal artifacts. Specifically, by adding the convolutional block attention module (CBAM) layers with a proper disentanglement parameter, experimental results confirm that we can get more improved MAR that preserves the detailed texture of the original image.

show abstract

Unpaired low‐dose computed tomography image denoising using a progressive cyclical convolutional neural network

et al. 2023

Medical Physics

View full text Add to dashboard Cite

BackgroundReducing the radiation dose from computed tomography (CT) can significantly reduce the radiation risk to patients. However, low‐dose CT (LDCT) suffers from severe and complex noise interference that affects subsequent diagnosis and analysis. Recently, deep learning‐based methods have shown superior performance in LDCT image‐denoising tasks. However, most methods require many normal‐dose and low‐dose CT image pairs, which are difficult to obtain in clinical applications. Unsupervised methods, on the other hand, are more general.PurposeDeep learning methods based on GAN networks have been widely used for unsupervised LDCT denoising, but the additional memory requirements of the model also hinder its further clinical application. To this end, we propose a simpler multi‐stage denoising framework trained using unpaired data, the progressive cyclical convolutional neural network (PCCNN), which can remove the noise from CT images in latent space.MethodsOur proposed PCCNN introduces a noise transfer model that transfers noise from LDCT to normal‐dose CT (NDCT), denoised CT images generated from unpaired CT images, and noisy CT images. The denoising framework also contains a progressive module that effectively removes noise through multi‐stage wavelet transforms without sacrificing high‐frequency components such as edges and details.ResultsCompared with seven LDCT denoising algorithms, we perform a quantitative and qualitative evaluation of the experimental results and perform ablation experiments on each network module and loss function. On the AAPM dataset, compared with the contrasted unsupervised methods, our denoising framework has excellent denoising performance increasing the peak signal‐to‐noise ratio (PSNR) from 29.622 to 30.671, and the structural similarity index (SSIM) was increased from 0.8544 to 0.9199. The PCCNN denoising results were relatively optimal and statistically significant. In the qualitative result comparison, PCCNN without introducing additional blurring and artifacts, the resulting image has higher resolution and complete detail preservation, and the overall structural texture of the image is closer to NDCT. In visual assessments, PCCNN achieves a relatively balanced result in noise suppression, contrast retention, and lesion discrimination.ConclusionsExtensive experimental validation shows that our scheme achieves reconstruction results comparable to supervised learning methods and has performed well in image quality and medical diagnostic acceptability.

show abstract

Unsupervised Denoising for Satellite Imagery using Wavelet Subband CycleGAN

Cited by 3 publications

References 33 publications

AdaIN-Based Tunable CycleGAN for Efficient Unsupervised Low-Dose CT Denoising

AdaIN-Based Tunable CycleGAN for Efficient Unsupervised Low-Dose CT Denoising

Unsupervised CT Metal Artifact Learning using Attention-guided beta-CycleGAN

Unpaired low‐dose computed tomography image denoising using a progressive cyclical convolutional neural network

Contact Info

Product

Resources

About