Dual-branch deep image prior for image denoising

Xu, Shaoping; Cheng, Xiaohui; Luo, Jie; Xiao, Nan; Xiong, Minghai; Zhou, Changfei

doi:10.1016/j.jvcir.2023.103821

Cited by 2 publications

(3 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…DBDIP [24] The Dual-branch Deep Image Prior model has a higher implementation complexity compared to the original DIP method, requiring more computational resources due to its dual-branch structure. The running time of the DBDIP method is longer than the original DIP method because of its iterative nature and more complex architecture.…”

Section: Methods Limitationmentioning

confidence: 99%

See 1 more Smart Citation

Enhancing Image Clarity with the Combined Use of REDNet and Attention Channel Module

Halim,

Putra Kusuma

2024

IJCDS

View full text Add to dashboard Cite

The primary aim of our study is to improve the efficacy of image denoising, specifically in situations when there is a limited availability of data, such as the BSD68 dataset. Insufficient data presents a challenge in achieving optimal outcomes due to the complexity involved in constructing models. In order to tackle this difficulty, we provide a method that incorporates Channel Attention, Batch Normalization, and Dropout approaches into the current REDNet framework. Our investigation indicates enhancements in performance parameters, such as PSNR (Peak Signal to Noise Ratio) and SSIM (Structural Similarity Index), across various levels of noise. With a noise level of 15, we obtained a Peak Signal-to-Noise Ratio (PSNR) of 34.9858 dB and a Structural Similarity Index (SSIM) of 0.9371. At a noise level of 25, our tests yielded a PSNR of 31.7886 decibels and an SSIM of 0.8876. In addition, at a noise level of 50, we achieved a Peak Signal-to-Noise Ratio (PSNR) of 27.9063 decibels and a Structural Similarity Index (SSIM) of 0.7754. The incorporation of Channel Attention, Batch Normalization, and Dropout has been demonstrated to be a crucial element in enhancing the efficacy of image denoising. The Channel Attention approach enables the model to choose and concentrate on crucial information inside the image, while Batch Normalization and Dropout techniques provide stability and mitigate overfitting issues throughout the training process. Our research highlights the effectiveness of these three strategies and emphasizes their integration as a novel way to address the constraints presented by the scarcity of data in image denoising jobs. This emphasizes the significant potential in creating dependable and effective image denoising methods when dealing with circumstances when there is a limited dataset.

show abstract

Section: Methods Limitationmentioning

confidence: 99%

“…DBDIP [24] method presents a new technique for reducing noise in images by utilizing additional information from two fundamental images created using a Y-Net structure. At first, the model divides the contracting path of a U-Net network into two branches to separately extract features using convolution and pooling techniques.…”

Section: Related Workmentioning

confidence: 99%

Enhancing Image Clarity with the Combined Use of REDNet and Attention Channel Module

Halim,

Putra Kusuma

2024

IJCDS

View full text Add to dashboard Cite

show abstract

“…All the processes, such as identification, inspection, and combination of images obtained with the help of cameras, are called image processing [14], [41]. Image processing applications have significantly developed in recent years [37], [80]. It is widely used in many fields, especially in defense, security, industry, law, health, and agriculture [20], [23], [42], [48], [73], [89], [91].…”

Section: Image Processingmentioning

confidence: 99%

Stock Detection of Iron and Steel Products with Image Processing and SDSS Decision Support System

2024

jjmie

View full text Add to dashboard Cite

The iron and steel industry is one of the leading sectors contributing to the economic power of countries. It is incorporated into numerous fields, including automotive, construction, manufacturing, agriculture, defense, and healthcare. Today, the increasing supply and demand balance in iron and steel products makes the stock management and control of the products crucial. The large quantity and variety of products in the iron and steel industry make stock management and control difficult. Creating stock for the business is a very costly process and is one of the important elements of the business. Successful stock management can make the business financially advantageous. This study aims to estimate the stock costs of different products and quantities according to different dates found in the business by using data mining. For this purpose, data mining classifier models are used, and estimated costs are found. By establishing a stock tracking system throughout the supply chain, the business should register all inventory movements of the products. It should work in an integrated way with stock management costing. Thus, unforeseen decreases and financial losses in products can be detected.The iron and steel sector, a key player in enhancing competition among nations and shaping the economic landscape, caters to the needs of various industries through its commitment to sustainable steel production. The industry has encountered challenges due to the swift advancements in recent years, impacting the costs associated with iron and steel products and posing issues for the sector. Tackling these challenges is essential for fostering national development and enhancing the competitiveness of businesses. This research delves into the elements influencing the expenses within the iron and steel industry, emphasizing the importance of minimizing factors leading to elevated costs. An automation system was developed using image processing techniques, and iron and steel products were analyzed. In the automation system, k-means and twostep, one of the clustering analysis methods, was applied.A decision support system defined as SDSS (Steel Decision Support System) and BIPS (Burak's Image Processing System)has been developed to determine the inventory costs of iron and steel products, adopting a data mining approach and including clustering methods to reveal similar inventory costs. Measurements with the SDSS-BIPS automation system were 94.4% successful. Withcost levels were determined by matching the products in the iron and steel business database with their costs.

show abstract

Dual-branch deep image prior for image denoising

Cited by 2 publications

References 22 publications

Enhancing Image Clarity with the Combined Use of REDNet and Attention Channel Module

Enhancing Image Clarity with the Combined Use of REDNet and Attention Channel Module

Stock Detection of Iron and Steel Products with Image Processing and SDSS Decision Support System

Contact Info

Product

Resources

About