Indoor occupants’ positions are significant for smart home service systems, which usually consist of robot service(s), appliance control and other intelligent applications. In this paper, an innovative localization method is proposed for tracking humans’ position in indoor environments based on passive infrared (PIR) sensors using an accessibility map and an A-star algorithm, aiming at providing intelligent services. First the accessibility map reflecting the visiting habits of the occupants is established through the integral training with indoor environments and other prior knowledge. Then the PIR sensors, which placement depends on the training results in the accessibility map, get the rough location information. For more precise positioning, the A-start algorithm is used to refine the localization, fused with the accessibility map and the PIR sensor data. Experiments were conducted in a mock apartment testbed. The ground truth data was obtained from an Opti-track system. The results demonstrate that the proposed method is able to track persons in a smart home environment and provide a solution for home robot localization.

The goal of this paper is to localize a resident in indoor environments by using motion information from distributed environmental sensors and body activity information from wearable sensors. The passive infrared sensor nodes distributed in a home provide binary information about human motion in their field of views, while the wearable inertial measurement unit sensor node collects motion data that can be used in body activity recognition, walking velocity, and heading estimation. Basic human activities such as sitting, sleeping, standing, and walking are recognized. We proposed a particle filter-based sensor fusion algorithm that takes advantage of the human location/activity correlation in indoor environments to increase the localization accuracy. Experiments were conducted in a mock apartment testbed. We used the ground truth data obtained from a motion capture system to evaluate the results.

Computer-aided automatic segmentation of retinal blood vessels plays an important role in the diagnosis of diseases such as diabetes, glaucoma, and macular degeneration. In this paper, we propose a multi-scale feature fusion retinal vessel segmentation model based on U-Net, named MSFFU-Net. The model introduces the inception structure into the multi-scale feature extraction encoder part, and the max-pooling index is applied during the upsampling process in the feature fusion decoder of an improved network. The skip layer connection is used to transfer each set of feature maps generated on the encoder path to the corresponding feature maps on the decoder path. Moreover, a cost-sensitive loss function based on the Dice coefficient and cross-entropy is designed. Four transformations—rotating, mirroring, shifting and cropping—are used as data augmentation strategies, and the CLAHE algorithm is applied to image preprocessing. The proposed framework is tested and trained on DRIVE and STARE, and sensitivity (Sen), specificity (Spe), accuracy (Acc), and area under curve (AUC) are adopted as the evaluation metrics. Detailed comparisons with U-Net model, at last, it verifies the effectiveness and robustness of the proposed model. The Sen of 0.7762 and 0.7721, Spe of 0.9835 and 0.9885, Acc of 0.9694 and 0.9537 and AUC value of 0.9790 and 0.9680 were achieved on DRIVE and STARE databases, respectively. Results are also compared to other state-of-the-art methods, demonstrating that the performance of the proposed method is superior to that of other methods and showing its competitive results.

Imbalanced voltage sharing during the turn-off transient is a challenge for series-connected silicon carbide (SiC) MOSFET application. This article first discusses the influence of the gate-drain discharge deviation on the voltage imbalance ratio, and its primary causes are also presented and verified by LTspice simulation. Accordingly, a novel active gate drive, which aims to compensate the discharge difference between devices connected in series, is proposed and analyzed. By only using the original output of the driving IC, the proposed gate drive is realized by implementing an auxiliary circuit on the existing commercial gate drive. Therefore, unlike other active gate drives for balancing control, no extra isolations for power/signal are needed, and the number of the devices in series is unlimited. The auxiliary circuit includes three subcircuits as a high-bandwidth current sink for regulating switching performance, a relative low-frequency but reliable sampling and control circuit for closed-loop control, and a trigger combining the former and the latter. The operational principle and the design guideline for each part are presented in detail. Experimental results validate the performance of the proposed gate drive and its voltage balancing control algorithm.

The new coronavirus, which has become a global pandemic, has confirmed more than 88 million cases worldwide since the first case was recorded in December 2019, causing over 1.9 million deaths. Since COIVD-19 lesions have clear imaging features on CT images, it is suitable for the auxiliary diagnosis and treatment of COVID-19. Deep learning can be used to segment the lesions areas of COVID-19 in CT images to help monitor the epidemic situation. In this paper, we propose a multi-point supervision network (MPS-Net) for segmentation of COVID-19 lung infection CT image lesions to solve the problem of a variety of lesion shapes and areas. A multi-scale feature extraction structure, a sieve connection structure (SC), a multi-scale input structure and a multi-point supervised training structure were implemented into MPS-Net. In order to increase the ability to segment various lesion areas of different sizes, the multi-scale feature extraction structure and the sieve connection structure will use different sizes of receptive fields to extract feature maps of various scales. The multi-scale input structure is used to minimize the edge loss caused by the convolution process. In order to improve the accuracy of segmentation, we propose a multi-point supervision training structure to extract supervision signals from different up-sampling points on the network. Experimental results showed that the dice similarity coefficient (DSC), sensitivity, specificity and IOU of the segmentation results of our model are 0.8325, 0.8406, 09988 and 0.742, respectively. The experimental results demonstrated that the network proposed in this paper can effectively segment COVID-19 infection on CT images. It can be used to assist the diagnosis and treatment of new coronary pneumonia.

Optical magnetic field sensors have gained prominence due to their compact size, high sensitivity, and broad dynamic range. Photonic crystal fiber (PCF) sensors have emerged as a significant role in optical sensing. Traditional numerical design methods of PCF such as finite element method (FEM) are computationally demanding and often restrict the design space due to iterative trial-and-error processes. In this work, a bidirectional design approach using a deep learning model based on Deep Neural Networks (DNNs) and Augmented Grey Wolf Optimizer (AGWO) is proposed for surface plasmon resonance (SPR)-based PCF magnetic field sensor. This bidirectional design method contains forward modeling and inverse design. The forward modeling predicts the optical responses of sensor structures accurately in 55.99 milliseconds with an R-squared value of 0.9942. The inverse design approach, with a tandem network configuration for addressing the non-uniqueness challenge in inverse design, identifies the sensor design aligning with desired optical responses in less than 0.5 seconds. In addition, AGWO is utilized to further enhance the accuracy and convergence ability of DNN. The key results indicate that our proposed approach reduces computational effort and enhances design effectiveness in PCF sensor design, which can also offer an alternative method for the design of various optical devices.