Image denoising is a challenging task that is essential in numerous computer vision and image processing problems. This study proposes and applies a generative adversarial network-based image denoising training architecture to multiple-level Gaussian image denoising tasks. Convolutional neural network-based denoising approaches come across a blurriness issue that produces denoised images blurry on texture details. To resolve the blurriness issue, we first performed a theoretical study of the cause of the problem. Subsequently, we proposed an adversarial Gaussian denoiser network, which uses the generative adversarial network-based adversarial learning process for image denoising tasks. This framework resolves the blurriness problem by encouraging the denoiser network to find the distribution of sharp noise-free images instead of blurry images. Experimental results demonstrate that the proposed framework can effectively resolve the blurriness problem and achieve significant denoising efficiency than the state-of-the-art denoising methods.

In the intelligent high-speed railway system, the security of deep neural networks-based high-speed train bogie fault diagnosis methods is challenged by adversarial attacks, which can mislead the model predictions with maliciously designed adversarial examples. However, existing methods do not consider the robustness against adversarial attacks. To address the aforementioned challenge, we propose a novel method called AdvSifter to perform robust fault diagnosis for the high-speed train bogie against adversarial attacks, which leverages adversarial training (AT) to guarantee the model with fundamental adversarial robustness. Besides, a defense algorithm called residual perturbation inversion (RPI) is developed to recover and remove the perturbations in adversarial examples to reduce the power of the adversarial examples. A defense module called SifterNet is designed to perform RPI to further improve the adversarial robustness of AdvSifter on the base of AT. Experimental results on a high-speed train bogie monitoring dataset demonstrate that our method outperforms state-of-the-art methods by a large margin.

Deep neural networks (DNNs) have been known to be vulnerable to adversarial attacks. Adversarial training (AT) is, so far, the only method that can guarantee the robustness of DNNs to adversarial attacks. However, the robustness generalization accuracy gain of AT is still far lower than the standard generalization accuracy of an undefended model, and there is known to be a trade-off between the standard generalization accuracy and the robustness generalization accuracy of an adversarially trained model. In order to improve the robustness generalization and the standard generalization performance trade-off of AT, we propose a novel defense algorithm called Between-Class Adversarial Training (BCAT) that combines Between-Class learning (BC-learning) with standard AT. Specifically, BCAT mixes two adversarial examples from different classes and uses the mixed between-class adversarial examples to train a model instead of original adversarial examples during AT. We further propose BCAT+ which adopts a more powerful mixing method. BCAT and BCAT+ impose effective regularization on the feature distribution of adversarial examples to enlarge between-class distance, thus improving the robustness generalization and the standard generalization performance of AT. The proposed algorithms do not introduce any hyperparameters into standard AT; therefore, the process of hyperparameters searching can be avoided. We evaluate the proposed algorithms under both white-box attacks and black-box attacks using a spectrum of perturbation values on CIFAR-10, CIFAR-100, and SVHN datasets. The research findings indicate that our algorithms achieve better global robustness generalization performance than the state-of-the-art adversarial defense methods.

Image-to-image conversion based on deep learning techniques is a topic of interest in the fields of robotics and computer vision. A series of typical tasks, such as applying semantic labels to building photos, edges to photos, and raining to de-raining, can be seen as paired image-to-image conversion problems. In such problems, the image generation network learns from the information in the form of input images. The input images and the corresponding targeted images must share the same basic structure to perfectly generate target-oriented output images. However, the shared basic structure between paired images is not as ideal as assumed, which can significantly affect the output of the generating model. Therefore, we propose a novel Input-Perceptual and Reconstruction Adversarial Network (IP-RAN) as an all-purpose framework for imperfect paired image-to-image conversion problems. We demonstrate, through the experimental results, that our IP-RAN method significantly outperforms the current state-of-the-art techniques.