The advancement of Internet Things (IoT) technologies leads to a wide penetration and large-scale deployment IoT systems across an entire city or even country. While are capable providing intelligent services, the large amount data collected processed in also raises serious security concerns. Many research efforts have been devoted design network intrusion detection system (NIDS) prevent misuse smart applications. However, existing approaches may suffer from issue limited imbalanced attack...

Nowadays, the concept of Internet Everything (IoE) is becoming a hotly discussed topic, which playing an increasingly indispensable role in modern intelligent applications. These applications are known for their real-time requirements under limited network and computing resources, thus it becomes highly demanding task to transform compute tremendous amount raw data cloud center. The edge–cloud infrastructure allows large be processed on nearby edge nodes then only extracted encrypted key...

Lung cancer is a leading cause of death worldwide, and its early detection usually performed with low-dose computed tomography. For lesions suspected abnormality by CT examination, cytological diagnosis lung cells collected biopsy first performed. However, atypical are the major challenge in malignant cell classification for cytotechnologists cytopathologists. In this study, we aimed to automatize from microscopic images using deep convolutional neural network (DCNN). Cytological specimens...

Physics theory integrated machine learning models enhance the interpretability and performance of artificial intelligence (AI) techniques to real-world industrial applications, such as fault detection diagnosis (FDD) air handling units (AHU). Traditional learning-based automated FDD model demonstrates a high classification accuracy with sufficient training data samples, however, suffers from physical interpretation models. In this article, Wasserstain generative adversarial network (WGAN) is...

Deep learning has been widely used for plant disease recognition in smart agriculture and proven to be a powerful tool image classification pattern recognition. However, it limited interpretability deep features. With the transfer of expert knowledge, handcrafted features provide new way personalized diagnosis diseases. irrelevant redundant lead high dimensionality. In this study, we proposed swarm intelligence algorithm feature selection [salp (SSAFS)] image-based detection. SSAFS is...

With the advantages of wide range, constant observation ability, and an active imaging mechanism, synthetic aperture radar (SAR) has been a preferrable choice for ship detection in complicated scenarios. However, existing algorithms, especially convolutional neural network (CNN), cannot achieve satisfactory accuracy generalization ability. Moreover, complex architectures limit their real-time performances on embedding or edge computing platforms. To handle these issues, this article proposes...

Insomnia is a common public health problem and an open biomedical research topic. results in various problems, including memory decline, decreases concentration weakens problem-solving ability. The insufficient sleep also leads to skin ageing, heart disease, high blood pressure, arrhythmia stroke. While it remains as global concern, quality improvement using modern technologies, such machine learning, classification virtual reality (VR), becomes hot problem. These technologies offer new...

As a high mortality disease, cancer seriously affects people's life and well-being. Reliance on pathologists to assess disease progression from pathological images is inaccurate burdensome. Computer aided diagnosis (CAD) system can effectively assist make more credible decisions. However, large number of labeled medical that contribute improve the accuracy machine learning algorithm, especially for deep in CAD, are difficult collect. Therefore, this work, an improved few-shot method proposed...

The energy storage capacity of electric vehicles (EVs) enables them to have the potential form an uninterruptible power system (UPS) when supply is interrupted in community. To solve problems related reasonable distribution and accurate voltage tracking control for EVs accessed community UPS, integrated cooperative strategy based on multi-agent proposed this paper. First, improved <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML"...

Various kinds of sensing data can be acquired for smart fault detection. Each signal source has spatial attributes and strong correlations exist between different sources. However, most the existing detection models are established in vector domain, which would destroy structure information embedded within multisource data. Besides, nonideal data, especially strong-noise mislabeled will seriously affect performance. Therefore, a matrix-form one-class model called robustness support matrix...

The accuracy and real-time of velocity measurement impact on the performance motor servo system. Incremental optical encoder measuring have advantages high easy-to miniaturization. typical method for measurements by is hybrid M / T method. poor in real time. Raised alternate M/T measurement. Design software hardware experiment results show feasibility correctness system, improves system with photoelectric

Multi-image hiding, which embeds multiple secret images into a cover image and is able to recover these with high quality, has gradually become research hotspot in the field of steganography. However, due need embed large amount data limited space, issues such as contour shadowing or color distortion often arise, posing significant challenges for multi-image hiding. In this paper, we propose StegaINR4MIH, novel implicit neural representation steganography framework that enables hiding within...

In view of the large and high real-time data transmission in current strapdown INS, USB interface was integrated into a inertial navigation processing system composed DSP, FPGA designed. The composition working principle were described; overall hardware design software introduced, where, function characteristics method based on FT245 focused on. is concision easy to realize, its cost low size small. Many experiments verified that can effectively accomplish solution flight posture...

The natural distribution of monitoring data is imbalanced, which has a negative impact on the training intelligent diagnosis models. Although researchers have proposed data-level and algorithm-level methods to solve this problem, these are only applicable small imbalance scenarios. In order correct anomalies model under large scenarios, paper proposes gradient harmonized loss that coordinates gradients each class prevent majority in imbalanced from dominating training. coordination based...