The Vision Transformer (ViT) architecture has been remarkably successful in image restoration. For a while, Convolutional Neural Networks (CNN) predominated in most computer vision tasks. Now, both CNN and ViT are efficient approaches that demonstrate powerful capabilities to restore a better version of an image given in a low-quality format. In this study, the efficiency of ViT in image restoration is studied extensively. The ViT architectures are classified for every task of image restoration. Seven image restoration tasks are considered: Image Super-Resolution, Image Denoising, General Image Enhancement, JPEG Compression Artifact Reduction, Image Deblurring, Removing Adverse Weather Conditions, and Image Dehazing. The outcomes, the advantages, the limitations, and the possible areas for future research are detailed. Overall, it is noted that incorporating ViT in the new architectures for image restoration is becoming a rule. This is due to some advantages compared to CNN, such as better efficiency, especially when more data are fed to the network, robustness in feature extraction, and a better feature learning approach that sees better the variances and characteristics of the input. Nevertheless, some drawbacks exist, such as the need for more data to show the benefits of ViT over CNN, the increased computational cost due to the complexity of the self-attention block, a more challenging training process, and the lack of interpretability. These drawbacks represent the future research direction that should be targeted to increase the efficiency of ViT in the image restoration domain.

High-quality images acquired from medical devices can be utilized to aid diagnosis and detection of various diseases. However, such images can be very expensive to acquire and difficult to store, and the process of diagnosis can consume significant time. Automatic diagnosis based on artificial intelligence (AI) techniques can contribute significantly to overcoming the cost and time issues. Pre-trained deep learning models can present an effective solution to medical image classification. In this paper, we propose two such models, ResNext101_32x8d and VGG19 to classify two types of brain tumor: pituitary and glioma The proposed models are applied to a dataset consisting of 1,800 MRI images comprising in two classes of diagnoses; glioma tumor and pituitary tumor. A single-image super-resolution (SISR) technique is applied to the MRI images to clarify and enhance their basic features, enabling the proposed models to enhance particular aspects of the MRI images and assist the training process of the models. These models are implemented using PyTorch and TensorFlow frameworks with hyper-parameter tuning, and data augmentation. Experimentally, receiver operating characteristic curve (ROCC), the error matrix, Precision and Recall are used to analyze the performance of the proposed model. Results obtained demonstrate that VGG19 and ResNext101_32x8d achieved testing accuracies of 99.98% and 100%, and loss rates of 0.0120 and 0.108, respectively. The F1-score, Precision, Recall, and the area under the ROC for VGG19 were 99.89%, 99.90%, 99.89%, and 100%, respectively, while for the ResNext101_32x8d they were all 100%. The proposed models when applied to MRI images to provide a quick and accurate approach to distinguishing between patients with pituitary and glioma tumors, and could aid doctors and radiologists in the screening of patients with brain tumors.

Remote Sensing (RS) images are usually captured at resolutions lower than those required. Deep Learning (DL)-based super-resolution (SR) architectures are typically used to increase the resolution artificially. In this study, we designed a new architecture called TESR (Two-stage approach for Enhancement and super-resolution), leveraging the power of Vision Transformers (ViT) and the Diffusion Model (DM) to increase the resolution of RS images artificially. The first stage is the ViT-based model, which serves to increase resolution. The second stage is an iterative DM pre-trained on a larger dataset, which serves to increase image quality. Every stage is trained separately on the given task using a separate dataset. The self-attention mechanism of the ViT helps the first stage generate global and contextual details. The iterative Diffusion Model helps the second stage enhance the image’s quality and generate consistent and harmonic fine details. We found that TESR outperforms state-of-the-art architectures on super-resolution of remote sensing images on the UCMerced benchmark dataset. Considering the PSNR/SSIM metrics, TESR improves SR image quality as compared to state-of-the-art techniques from 34.03/0.9301 to 35.367/0.9449 in the scale ×2. On a scale of ×3, it improves from 29.92/0.8408 to 32.311/0.91143. On a scale of ×4, it improves from 27.77/0.7630 to 31.951/0.90456. We also found that the Charbonnier loss outperformed other loss functions in the training of both stages of TESR. The improvement was by a margin of 21.5%/14.3%, in the PSNR/SSIM, respectively. The source code of TESR is open to the community.