With the purpose of further mastering and grasping course speech signal processing, a novel Android-based, mobile-assisted educational platform (AEPS) is proposed in this paper. The goal work was to design AEPS as an signal-processing auxiliary system by simulating analysis methods commonly used processing bridging gap for transition from undergraduate study industry practice or academic research. presented highly intuitive, easy-to-interpret strongly maneuverable graphical user interface....

Object detection is a crucial topic in computer vision. Mask Region‐Convolution Neural Network (R‐CNN) based methods, wherein large intersection over union (IoU) threshold chosen for high quality samples, have often been employed object detection. However, the performance of such methods deteriorates when samples are reduced. To address this, authors propose an improved R‐CNN‐based method: ResNet Group Cascade (RGC) R‐CNN. First, they compared with different layers, finding that...

Millimeter wave radar as a type of noncontact sensor can more conveniently and insensibly obtain breathing heartbeat signals. However, due to the weak echo signal influence noise, there may be phase ambiguity separation distortion signals during processing, which affects accuracy detection. In this paper, an enhanced extended differential cross multiplication (EDACM) algorithm is presented chest position without ambiguity. A specific order finite impulse response (FIR) filter with Blackman...

The traditional K-singular value decomposition (K-SVD) algorithm has poor image-denoising performance under strong noise. An is proposed based on improved K-SVD and dictionary atom optimization. First, a correlation coefficient-matching criterion used to obtain sparser representation of the image dictionary. noise detected according structural complexity intensity removed optimize Then, non-local regularity incorporated into denoising model further improve performance. Results simulated...

The automated classification of breast cancer histopathological images is one the important tasks in computer-aided diagnosis systems (CADs). Due to characteristics small inter-class and large intra-class variances images, extracting features for difficult. To address this problem, an improved autoencoder (AE) network using a Siamese framework that can learn effective from CAD was designed. First, inputted image processed at multiple scales Gaussian pyramid obtain multi-scale features....

Nonuniform aperture synthesis radiometers (NASRs) have the advantage of flexibility in practical applications. However, main limitation NASRs for Earth observation is poor reconstruction quality. In this letter, array factor 1-D NASR analyzed. Array forming (AFF) introduced to improve quality NASRs. The limitations AFF on configurations are given. that meets given called optimizable (ONASR). numerical results and experimental demonstrate able ONASRs. To assess performance thoroughly,...

Battery failure has traditionally been a major concern for electric vehicle (EV) safety, and early fault diagnosis will reduce many EV safety accidents. However, the short-circuit signal is generally very weak, so it still challenge to achieve timely warning of battery failure. In this paper, an initial microfault method proposed data vehicles in actual operation. First, robust locally weighted regression smoothing that can effectively remove noisy retain characteristics. Second,...

Abstract With the rapid development of global electric vehicles (EVs) market, accurately predicting charging times is significant importance for promoting widespread adoption EVs and enhancing efficiency infrastructure. Existing prediction methods often disregard battery aging predominantly use single-model approaches, resulting in limited predictive accuracy. This paper proposes a multi-model fusion-based method times. The approach utilizes data from 10 across various regions operational...

Complexities in processing human motion are possessed by lower limb exoskeletons. In this paper, a multi-task recognition model, IPTGNet, is proposed for the locomotion modes. Temporal convolutional network and gated recurrent unit parallelly fused through dynamic tuning of hyperparameters using improved particle swarm optimization algorithm. The experimental results demonstrate that faster more stable convergence achieved IPTGNet with rate 99.47% standard deviation 0.42%. Furthermore,...

It is quite simple for foreign objects to attach themselves transmission line corridors because of the wide variety laying and complex, changing environment. If these are not found removed in a timely manner, they can have significant impact on lines’ ability operate safely. Due problem poor accuracy object identification image inspection, we provide an improved YOLOX technique detection lines. The method improves target network by first using Atrous Spatial Pyramid Pooling increase...

Abstract Nuclei instance segmentation is an important task in medical image analysis involving cell‐level pathological analysis, which of great significance for many biomedical applications. a challenging due to edge adhesions and the distribution numerous tiny dense nuclei. In this work, nuclei framework, namely, improved BlendMask proposed. order improve performance detection small objects adhering nuclei, two components, including dilated convolution aggregation module (DCA) context...

In the field of inter-satellite laser communication, achieving high-quality communication and compensating for Doppler frequency shift caused by relative motion necessitate lasers with narrow linewidths, low phase noise, ability to achieve mode-hop-free tuning within a specific range. To this end, paper investigates novel external cavity diode (ECDL) frequency-selective F-P etalon structure, leveraging structure in conjunction an auxiliary filter single longitudinal mode selection. The...

Locomotion mode recognition in humans is fundamental for flexible control wearable-powered exoskeleton robots. This article proposes a hybrid model that combines dense convolutional network (DenseNet) and long short-term memory (LSTM) with channel attention mechanism (SENet) locomotion recognition. DenseNet can automatically extract deep-level features from data, while LSTM effectively captures long-dependent information time series. To evaluate the validity of model, inertial measurement...

Abstract Deep learning techniques have achieved specific results in recording device source identification. The features include spatial information and certain temporal information. However, most identification methods based on deep only use representation from features, which cannot make full of Therefore, this paper, to fully explore the source, we propose a new method for fusion feature by using an end-to-end framework. From perspective, designed two kinds networks extract Afterward,...

The global demand for electric power has been greatly increasing because of industrial development and the change in people’s daily life. A lot overhead transmission lines have installed to provide reliable across long distancess. Therefore, research on inspection is very important preventing sudden wide-area outages. In this paper, we propose an Overhead Transmission Line Classifier (OTL-Classifier) based deep learning techniques classify images returned by future unmanned maintenance...

A novel Compressed-Sensing-based (CS-based) Distributed Video Coding (DVC) system, called Adaptive Compressed Sensing (DISACOS), is proposed in this paper. In the input frames are divided into key and non-key frames, which encoded by block CS sampling. The as measurements at substantially higher rates than decoded Smoothed Projected Landweber (SPL) algorithm using multi-hypothesis predictions. For a small number of first transmitted to detect blocks having low-quality Side Information (SI)...