Image denoising using deep convolutional autoencoder with feature pyramids

Çetinkaya, Ekrem; Kıraç, Mustafa

doi:10.3906/elk-1911-138

Cited by 4 publications

(7 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The parameters used during the training process of the autoencoder can be seen in Table 1. [4][5][6][7][8][9][10][11] are the used dataset, similarity metrics and parameters. This study differs from the similar studies in the literature by testing well-known traditional and deep learning-based noise removal methods from face images of well-known noises with different metrics.…”

Section: Figure 9 the Proposed Cdae Model For Image Denoisingmentioning

confidence: 99%

See 1 more Smart Citation

A Comparative Study on Denoising from Facial Images Using Convolutional Autoencoder

Darici

Erdem

2023

Gazi University Journal of Science

View full text Add to dashboard Cite

Denoising is one of the most important preprocesses in image processing. Noises in images can prevent extracting some important information stored in images. Therefore, before some implementations such as image classification, segmentation, etc., image denoising is a necessity to obtain good results. The purpose of this study is to compare the deep learning techniques and traditional techniques on denoising facial images considering two different types of noise (Gaussian and Salt&Pepper). Gaussian, Median, and Mean filters have been specified as traditional methods. For deep learning methods, deep convolutional denoising autoencoders (CDAE) structured on three different optimizers have been proposed. Both accuracy metrics and computational times have been considered to evaluate the denoising performance of proposed autoencoders, and traditional methods. The utilized standard evaluation metrics are the peak signal to noise ratio (PSNR) and structural similarity index measure (SSIM). It has been observed that overall, while the traditional methods gave results in shorter times in terms of computation times, the autoencoders performed better concerning the evaluation metrics. The CDAE based on the Adam optimizer has been shown the best results in terms of PSNR and SSIM metrics on removing both types of noise

show abstract

Section: Figure 9 the Proposed Cdae Model For Image Denoisingmentioning

confidence: 99%

“…There are various studies that use autoencoders as a noise reduction method [5][6][7][8]. [9] has created a convolutional autoencoder to construct noiseless medical images from noisy medical images.…”

Section: Introductionmentioning

confidence: 99%

A Comparative Study on Denoising from Facial Images Using Convolutional Autoencoder

Darici

Erdem

2023

Gazi University Journal of Science

View full text Add to dashboard Cite

show abstract

“…In digital image processing and computer vision, image denoising is considered an ill-posed inverse problem [15][16][17][18][19][20]. Image denoising aims to remove the noise from an image and restore a latent clean image.…”

Section: Cnn-based Approaches For Image Denoising/despecklingmentioning

confidence: 99%

“…The first stage employs noise estimation using channel attention; in the second stage, multiscale features are extracted, and then feature fusion is performed using kernel selecting operation. In [15], to tackle the problem of various noise levels, a CNN auto-encoder network with a feature pyramid is proposed for additive white Gaussian noise (AWGN) suppression. A pyramid-aware network was proposed in [32] for blurry image restoration.…”

Section: Pyramid Networkmentioning

confidence: 99%

PAN-DeSpeck: A Lightweight Pyramid and Attention-Based Network for SAR Image Despeckling

Yasmeen,

Usman Yaseen,

Sohaib Ali

et al. 2023

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

SAR images commonly suffer from speckle noise, posing a significant challenge in their analysis and interpretation. Existing convolutional neural network (CNN) based despeckling methods have shown great performance in removing speckle noise. However, these CNN-based methods have a few limitations. They do not decouple complex background information in a multi-resolution manner. Moreover, they have deep network structures that may result in many parameters, limiting their applicability to mobile devices. Furthermore, extracting key speckle information in the presence of complex background is also a major problem with SAR. The proposed study addresses these limitations by introducing a lightweight pyramid and attention-based despeckling (PAN-Despeck) network. The primary objective is to enhance image quality and enable improved information interpretation, particularly on mobile devices and scenarios involving complex backgrounds. The PAN-Despeck network leverages domainspecific knowledge and integrates Gaussian Laplacian image pyramid decomposition for multi-resolution image analysis. By utilizing this approach, complex background information can be effectively decoupled, leading to enhanced despeckling performance. Furthermore, the attention mechanism selectively focuses on key speckle features and facilitates complex background removal. The network incorporates recursive and residual blocks to ensure computational efficiency and accelerate training speed, making it lightweight while maintaining high performance. Through comprehensive evaluations, it is demonstrated that PAN-Despeck outperforms existing image restoration methods. With an impressive average peak signal-to-noise ratio (PSNR) of 28.355114 and a remarkable structural similarity index (SSIM) of 0.905467, it demonstrates exceptional performance in effectively reducing speckle noise in SAR images. The source code for the PAN-DeSpeck network is available on GitHub.

show abstract

“…However, in several studies discussed about the accuracy of visual inspection alone addressing the high error rates it has been highlighted that diagnosis can be missed if visual inspection is used on its own [10]. It has been also pointed out that apart from improving the expertise of clinicians, this field should benefit more from new technologies and tools [11][12]. In this context, deep learning based computer aided (DL-CAD) systems seem to be a great candidate for improving the diagnostic accuracy.…”

Section: Introductionmentioning

confidence: 99%

Cilt Kanseri Teşhisi için Konvolüsyonel Sinir Ağları Tabanlı Bilgisayar Destekli Tanıda (CNN-CAD) Dijital Görüntü Kalitesinin İyileştirilmesi

YALÇIN

Gursel

2021

Çukurova Üniversitesi Mühendislik Fakültesi Dergisi

View full text Add to dashboard Cite

The practice of detecting skin cancer is based primarily on a visual examination by a dermatologist, followed by a series of tests for a more accurate diagnosis. The concept "the earlier cancer is detected in its natural history, the more effective the treatment is likely to be" is also valid for skin cancer. Hence, any delayed or missed diagnosis can lead to a more severe clinical stage or, what's worse, death. On the other hand, the lack of biomarkers in clinical use brings about overdiagnosis and unnecessary biopsies. DL-CAD system seems to be an excellent candidate for improving diagnostic accuracy and reducing unnecessary treatments. However, the vast majority of conventional CADs manipulate dermoscopic images, which require not only costly equipment but also time-consuming processing. Despite the difficulties with precision, state-of-the-art DL-CAD systems provide an interpretation using digital images, requiring no expertise in cost-effective dermoscopic image capture and interpretation. Preprocessing methods play a crucial role in solving this problem. This study presents results with regard to pre-processing steps to improve the images to be used in the diagnosis of the 5 most common skin cancer types for the proposed CNN based ResNet50 deep learning model. To the best of our knowledge it is the first time that ResNet50 deep-learning model has been utilized in diagnosis of skin cancer.

show abstract

Image denoising using deep convolutional autoencoder with feature pyramids

Cited by 4 publications

References 16 publications

A Comparative Study on Denoising from Facial Images Using Convolutional Autoencoder

A Comparative Study on Denoising from Facial Images Using Convolutional Autoencoder

PAN-DeSpeck: A Lightweight Pyramid and Attention-Based Network for SAR Image Despeckling

Cilt Kanseri Teşhisi için Konvolüsyonel Sinir Ağları Tabanlı Bilgisayar Destekli Tanıda (CNN-CAD) Dijital Görüntü Kalitesinin İyileştirilmesi

Contact Info

Product

Resources

About