Insect pests are one of the main threats to commercially important crops. An effective insect pest recognition method can avoid economic losses. In this paper, we proposed a new and simple structure based on original residual block named as feature reuse which combines from input signal with signal. each block, it enhances capacity representation by learning half feature. By stacking obtained network (FR-ResNet) evaluated performance IP102 benchmark dataset. The experimental results showed...

Insect pest control is considered as a significant factor in the yield of commercial crops. Thus, to avoid economic losses, we need valid method for insect recognition. In this paper, proposed feature fusion residual block perform recognition task. Based on original block, fused from previous layer between two 1×1 convolution layers signal branch improve capacity block. Furthermore, explored contribution each group model performance. We found that adding blocks earlier groups promotes...

Earlier insect pest recognition is one of the critical factors for agricultural yield. Thus, an effective method to recognize category pests has become significant issues in field. In this article, we proposed a new residual block learn multiscale representation. each block, it contains three branches: parameter-free, and others contain several successive convolution layers. Moreover, module embedded into recalibrate channelwise feature response model relationship branches. By stacking kind...

Laser additive manufacturing (LAM) technology is one of the most representative advanced technology, and critical currently a hot research at home abroad. This paper mainly introduces forming principle, characteristics status two typical LAM technologies--selective laser melting (SLM) cladding deposition (LCD) technology. Moreover, some application examples are listed, its future development trend discussed.

Accurate power load forecasting plays an important role in the dispatching and security of grid. In this paper, a mathematical model for based on random forest regression (RFR) was established. The input parameters RFR were determined by means grid search algorithm. prediction results compared with those several other common machine learning methods. It found that coefficient determination (R2) test set highest, reaching 0.514 while corresponding mean absolute error (MAE) squared (MSE)...

Abstract Short-term power load forecasting has always been a very important position in grid operation. Improving prediction accuracy great significance for improving the company’s dispatch efficiency and economic benefits. In view of feed-forward neural network cannot learn time series relationship between data, hybrid depth model based on improved bacterial foraging algorithm (IBFO) long short-term memory (LSTM) is proposed forecasting. Using historical data its associated temperature date...

Smart grid is willing to make full advantage of distributed clean energy alleviate crisis and environmental problems. However, renewable usually invisible uncontrollable for the current power system, there are intermittent problems in its generation. Therefore, how achieve balance, maintain safe operation, ensure reliability quality supply when reaches a high penetration huge challenge. Blockchain as one research hotspots brings about new solution approach dilemma. The two fields have many...

Deep convolutional neural networks (CNNs) have significantly enhanced image classification in the past decade. This paper proposes Multi-branch Cascade Receptive Field Residual Networks (MCRF-ResNets) based on original Network (ResNet) architecture for and object detection. MCRF-ResNets incorporate multiple branches with different receptive field (RF) sizes to improve performance. Each MCRF residual block contains a 5 × 3 RF-sized convolution block. The blocks recognize larger smaller...

Objective SWI image signal is related to venous reflux disorder and perfusion defect. Computed tomography (CTP) contains information in space time. There a complementary basis between them affect the prognosis of cerebral infarction. Methods Sixty-six patients included retrospective study were designated as training set. Effective indicator features imaging radiomic peri-infarction area on Susceptibility weighted (SWI) CTP modality images extracted from each case. Thirty-three prospectively...

A widely accepted Chinese emotional speech database with abundant spontaneous speeches is essential to recognition and affective computing. This paper presents a new method of constructing such audio-visual database. The source materials come from variety videos in Chinese. Voice Activity Detection technology introduced catch the sets start time end syntactic boundaries dialogue. times are helpful following extracting processing ensure reaching complete phrase or sentence. Microsoft Emotion...

Abstract We propose an Inner-Imaging three-dimensional (3D) attentional feature fusion module for a residual network, which is simple yet effective approach networks. In our attention module, we constructed 3D soft map to refine the input feature. The fuses features from different dimensions, including channel and spatial axes, create map. Then, implemented further fuse features. Lastly, outputs that applied branch. can also model relationship between dimensions achieve interaction This...

Aiming at the problem that online examination is difficult to monitor, a remote intelligent monitoring system for based on gaze direction classification proposed. In experimental environment, situation, dataset with nine types of directions multi-classification problems established, and image model used training. The same comparative experiments ConvNeXt network ResNet network, finally selected as backbone network. trained performs well in weighted accuracy recall rate test set, reaching...

As the power efficiency is affected by every aspect of a wind tur-bine and its blades, 2D-computational investigation several keyparameters proposed after which actual experiments are performed.Six parameters examined in computational investigation: instal-lation angle; blade airfoil; chord length; number blades; radius ofan impellor; speed. To find out optimum performance ofa Vertical Axis Wind Turbine or VAWT, one appropriate parameter isselected at each section simulation progression....