Multi-scale Xception based depthwise separable convolution for single image super-resolution

Abstract

The main target of Single image super-resolution is to recover high-quality or high-resolution image from degraded version of low-quality or low-resolution image. Recently, deep learning-based approaches have achieved significant performance in image super-resolution tasks. However, existing approaches related with image super-resolution fail to use the features information of low-resolution images as well as do not recover the hierarchical features for the final reconstruction purpose. In this research work, we have proposed a new architecture inspired by ResNet and Xception networks, which enable a significant drop in the number of network parameters and improve the processing speed to obtain the SR results. We are compared our proposed algorithm with existing state-of-the-art algorithms and confirmed the great ability to construct HR images with fine, rich, and sharp texture details as well as edges. The experimental results validate that our proposed approach has robust performance compared to other popular techniques related to accuracy, speed, and visual quality.

Feasibility of using AI to auto-catch responsible frames in ultrasound screening for breast cancer diagnosis

“…In our previous work, 44 we had completed training and testing four AI models (Xception, 45 MobileNet, 46 ResNet50, 47 and DenseNet121 48 ) with three input image resolutions (224 × 224, 320 × 320, and 448 × 448 pixels). According to the inclusion and exclusion criteria, 3447 females breast nodules were finally included.…”

A zero-shot fault attribute transfer learning method for compound fault diagnosis of power circuit breakers

“…Depthwise separable convolution. Depthwise separable convolution is a lightweight convolution structure that divides traditional convolution into channel by channel convolution and point by point convolution [31][32][33][34][35]. Channel dimension information and spatial dimension information are mapped separately to realize joint mapping of traditional convolution in two steps.…”

Spirits quality classification based on machine vision technology and expert knowledge

“…In this section, a total of 15 different CNN backbone models with different structures are utilized for classification, including five deep models (VGG16 [22], VGG19 [22], ResNet50 [23], InceptionV3 [24,25], DenseNet121 [26], Xception [27,28]) and nine lightweight models (SqueezeNet [29], ShuffleNet [30], NasNet_Mobile [31], MobileNet [30], NobileNetv2 [32], MobileNetv3_Small [33], MobileNetv3_Large [34], EfficientNetB0 [35], EfficientNetB3 [35]). All CNN models in this paper are trained using pre-trained weights to speed up the training process.…”