Transfer learningDomain adaptationWasserstein distanceConvolutional neural networksIntelligent fault diagnosisIntelligent fault diagnosis is one critical topic of maintenance solution for mechanical systems. Deep learning models, such as convolutional neural networks (CNNs), have been successfully applied to fault...
Transfer-based Adversarial Attack论文整理 对抗攻击分为白盒攻击和黑盒攻击,白盒攻击可以拿到模型的所有信息,几乎所有方法成功率都在90%以上,故不在此赘述,最近对抗攻击(Adversarial Attack)的趋势是黑盒攻击和物理攻击。黑盒攻击分为两种,一种为查询(Query-based)攻击,一种为迁移(Transfer-based)攻击,查询攻击通过...
Deep Transfer Learning-Based Intelligent Fault Diagnosis This chapter presents intelligent fault diagnosis methods based on deep transfer learning, which largely enhance the fault diagnosis performance in the rea... Y Lei,N Li,X Li 被引量: 0发表: 2023年 加载更多研究点推荐 Intelligent fault diagnosi...
Deep learning (DL) models are becoming pervasive and applicable to computer vision, image processing, and synthesis problems. The performance of these models is often improved through architectural configuration, tweaks, the use of enormous training data, and skillful selection of hyperparameters. The ...
Handcrafted and deep learning (DL) radiomics are popular techniques used to develop computed tomography (CT) imaging-based artificial intelligence models for COVID-19 research. However, contrast heterogeneity from real-world datasets may impair model per
Therefore, suffering from an imbalanced dataset, training a deep learning model for generalizing across samples to achieve precise RUL estimation proves to be a tedious task to solve. Thus, there is a need to alleviate the imbalanced dataset problem as it is a limiting factor to the overall ...
对抗样本论文学习(3):Practical Black-Box Attacks against Machine Learning Papernot等 通过用合成的数据集训练一个局部替代模型来模拟黑箱模型,其中数据集的标签由黑箱模型通过查询给出。然后通过代替模型来生成对抗样本来对黑盒模型进行攻击。 这些方法都是通过queries来获得黑盒模型的知识,然后通过训练代替模型来生成对...
Generative adversarial networks (GAN) [29], proposed by Ian Goodfellow, have revolutionized the deep learning field. GAN is one of the most widely used techniques for image generation. It generates fake data consistent with its distribution by continuously approximating the data distribution of real ...
Pixel blitting and geometrical transformations (orange and green in Fig.1a, respectively), both most used in the deep learning literature for their simplicity of implementation and increase in performance, remain the choice by excellence for natural (see Fig.4a–b in Karras et al.8) as well ...
A supervised deep learning model based on CNN called COVID-NET has been proposed in23 leading to 93.3% test accuracy, on a test set of 100 samples of normal, pneumonia, and Covid-19 chest X-ray (CXR) images from the COVIDx dataset24. All other images were used for training the ...