Results 31 to 40 of about 47,227 (224)

Deep Learning Reconstruction Enables Diagnostic-Quality 0.4T Knee and Spine MRI in One-Third of the Time. [PDF]

NMR Biomed
The deep learning model CIRIM successfully accelerated knee and spine data up to an acceleration factor of 3, after optimizing the undersampling mask and loss function. The model demonstrated robustness and generalizability to different contrasts, matrix sizes, orientations, and anatomies. ABSTRACT There has been a growing interest in low‐field MRI due
van den Berg DM, Varriale R, Ferrando F, Traverso P, Balbi L, Belgiorno G, Beekman K, Kox L, Strijkers GJ, Caan MWA.   +9 more
europepmc   +2 more sources

Optimizing Multiscale SSIM for Compression via MLDS [PDF]

IEEE Transactions on Image Processing, 2012
A crucial step in the assessment of an image compression method is the evaluation of the perceived quality of the compressed images. Typically, researchers ask observers to rate perceived image quality directly and use these rating measures, averaged across observers and images, to assess how image quality degrades with increasing compression.
Charrier, Christophe, Knoblauch, Kenneth, Maloney, Laurence, Bovik, Alan, Moorthy, Anush   +4 more
openaire   +4 more sources

A Transfer-Learning Approach for Accelerated MRI using Deep Neural Networks

, 2019
Purpose: Neural networks have received recent interest for reconstruction of undersampled MR acquisitions. Ideally network performance should be optimized by drawing the training and testing data from the same domain. In practice, however, large datasets
Dar, Salman Ul Hassan, Çatlı, Ahmet Burak, Çukur, Tolga, Özbey, Muzaffer   +3 more
core   +2 more sources

Perceptual Quality Study on Deep Learning based Image Compression

, 2019
Recently deep learning based image compression has made rapid advances with promising results based on objective quality metrics. However, a rigorous subjective quality evaluation on such compression schemes have rarely been reported.
Akyazi, Pinar, Cheng, Zhengxue, Ebrahimi, Touradj, Katto, Jiro, Sun, Heming   +4 more
core   +1 more source

Video quality assessment method: MD-SSIM [PDF]

2015 IEEE International Conference on Consumer Electronics - Taiwan, 2015
In this paper, video quality assessment (VQA) for compression losses is the main focus. A new method, MD-SSIM (Mean Squared Error Difference SSIM) is used for detecting the spatial distortion. For the temporal part, the differences of the SSIM scores of each frame are used to form the quality scores.
Woei-Tan Loh, David B. L. Bong
openaire   +1 more source

nOMP: A New Sparse Solution to Enhance the SSIM Levels of OMP-Based Encoded Images

IEEE Access, 2022
This work comprises the development of a quality enhancement technique for image encoders that use compressive sensing. The recommended solution seeks to maximize the perceptual quality based objective function, unlike other sparse representation ...
A. N. Omara, Hesham Farouk, Sayed A. Mohamed   +2 more
doaj   +1 more source

Learning to Generate Images with Perceptual Similarity Metrics

, 2017
Deep networks are increasingly being applied to problems involving image synthesis, e.g., generating images from textual descriptions and reconstructing an input image from a compact representation.
Liao, Renjie, Mozer, Michael C., Ridgeway, Karl, Roads, Brett D., Snell, Jake, Zemel, Richard S.   +5 more
core   +1 more source

Deep Multiple Description Coding by Learning Scalar Quantization

, 2019
In this paper, we propose a deep multiple description coding framework, whose quantizers are adaptively learned via the minimization of multiple description compressive loss. Firstly, our framework is built upon auto-encoder networks, which have multiple
Bai, Huihui, Wang, Anhong, Zhao, Lijun, Zhao, Yao   +3 more
core   +1 more source

Application of convolutional neural networks trained on optical images for object detection in radar images

Компьютерная оптика
Due to the small number of annotated radar image datasets, the use of optical images for training neural networks designed to detect objects in radar images seems promising.
V.A. Pavlov, A.A. Belov, S.V. Volvenko, A.V. Rashich   +3 more
doaj   +1 more source

A novel image augmentation based on statistical shape and intensity models: application to the segmentation of hip bones from CT images

European Radiology Experimental, 2023
Background The collection and annotation of medical images are hindered by data scarcity, privacy, and ethical reasons or limited resources, negatively affecting deep learning approaches.
Jérôme Schmid, Lazhari Assassi, Christophe Chênes   +2 more
doaj   +1 more source