Image Denoising With Generative Adversarial Networks and its Application to Cell Image Enhancement

Chen, Songkui; Shi, Daming; Sadiq, Muhammad; Cheng, Xiaochun

doi:10.1109/access.2020.2988284

Cited by 32 publications

(14 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For evaluation purposes, we compared our proposed method with the latest state-of-the-art image denoising approaches. The compared techniques included Dn-CNN [ 23 ], FFDNet [ 24 ], perceptually inspired denoising method [ 46 ], and ID-MSE-WGAN [ 45 ]. The Dn-CNN and ID-MSE-WGAN predicted the noise first and then constructed the target images by subtracting that learned noise from the input images.…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…Many solutions [ 45 , 46 , 59 ] to denoising problems utilized skip-connection in the denoiser network, transporting the data directly from the input to the output through the network for resolving the disappearing gradient issue. On one hand, skip-connections help resolve the vanishing gradient issue.…”

Section: Methodsmentioning

confidence: 99%

“…The deep-CNNs suffer from a disappearing gradient issue during the training process. Consequently, many previous studies [ 45 , 46 , 59 ] have utilized skip-connections in a generator network to easily allow a gradient to earlier network layers. Unfortunately, such skip-connections bring unwanted data straight from the input images to the constructed images, reducing the constructing images’ visual quality.…”

Section: Related Workmentioning

confidence: 99%

“…In the literature, most of the methods used residual learning for image denoising tasks. For example, based on Wasserstein generative adversarial networks (WGAN) [ 44 ], Chen et al [ 45 ] proposed an image denoising training scheme and used residual learning for the generator network. In the residual learning image denoising methods, the network learns the residual image (noise image) first and subtracts the residual image from the input image to get the clean image.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Adversarial Gaussian Denoiser for Multiple-Level Image Denoising

Khan

Jin

Haider

et al. 2021

Sensors

View full text Add to dashboard Cite

Image denoising is a challenging task that is essential in numerous computer vision and image processing problems. This study proposes and applies a generative adversarial network-based image denoising training architecture to multiple-level Gaussian image denoising tasks. Convolutional neural network-based denoising approaches come across a blurriness issue that produces denoised images blurry on texture details. To resolve the blurriness issue, we first performed a theoretical study of the cause of the problem. Subsequently, we proposed an adversarial Gaussian denoiser network, which uses the generative adversarial network-based adversarial learning process for image denoising tasks. This framework resolves the blurriness problem by encouraging the denoiser network to find the distribution of sharp noise-free images instead of blurry images. Experimental results demonstrate that the proposed framework can effectively resolve the blurriness problem and achieve significant denoising efficiency than the state-of-the-art denoising methods.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Adversarial Gaussian Denoiser for Multiple-Level Image Denoising

Khan

Jin

Haider

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…Dimitrakopoulos et al [26] proposed a new GAN-based model for data augmentation that is suitable for the simultaneous production of synthetic cell images with their segmentation maps. In addition, Chen et al [27] used WGAN to denoise cell images and obtained cell images with clear features, providing a certain practical basis for generating cell cycle images with WGAN-GP.…”

Section: Wgan-gpmentioning

confidence: 99%

An Imbalanced Image Classification Method for the Cell Cycle Phase

2021

View full text Add to dashboard Cite

The cell cycle is an important process in cellular life. In recent years, some image processing methods have been developed to determine the cell cycle stages of individual cells. However, in most of these methods, cells have to be segmented, and their features need to be extracted. During feature extraction, some important information may be lost, resulting in lower classification accuracy. Thus, we used a deep learning method to retain all cell features. In order to solve the problems surrounding insufficient numbers of original images and the imbalanced distribution of original images, we used the Wasserstein generative adversarial network-gradient penalty (WGAN-GP) for data augmentation. At the same time, a residual network (ResNet) was used for image classification. ResNet is one of the most used deep learning classification networks. The classification accuracy of cell cycle images was achieved more effectively with our method, reaching 83.88%. Compared with an accuracy of 79.40% in previous experiments, our accuracy increased by 4.48%. Another dataset was used to verify the effect of our model and, compared with the accuracy from previous results, our accuracy increased by 12.52%. The results showed that our new cell cycle image classification system based on WGAN-GP and ResNet is useful for the classification of imbalanced images. Moreover, our method could potentially solve the low classification accuracy in biomedical images caused by insufficient numbers of original images and the imbalanced distribution of original images.

show abstract