Combination of the Improved Diffraction Nonlocal Boundary Condition and Three-Dimensional Wide-Angle Parabolic Equation Decomposition Model for Predicting Radio Wave Propagation by Ruidong Wang, Guizhen Lu, Rongshu Zhang, Weizhang Xu

“…Diffraction nonlocal boundary condition (BC) is one kind of the transparent boundary condition which is used in the finite-difference (FD) parabolic equation (PE). The greatest advantage of the diffraction nonlocal boundary condition is that it can absorb the wave completely by using one layer of grid. …”
Get full text

INFLUENCE OF THE AVERAGE WEIGHTED ESTIMATION TYPE ON THE DEPENDENCE OF THE COMPLEX QUALITY INDEX ON THE PARAMETERS OF OBJECT by A. M. Dolzhanskiy, O. A. Bondarenko, Ye. A. Petlyovaniy

“…It includes single quality indicators with their significance factors. The convolution of the corresponding dependencies represents average weighted quantities: arithmetic, geometric, harmonic, quadratic, etc. …”
Get full text

Adaptive pixel attention network for hyperspectral image classification by Yuefeng Zhao, Chengmin Zai, Nannan Hu, Lu Shi, Xue Zhou, Jingqi Sun

“…More importantly, we also propose a new Adaptive Pixel Attention mechanism, which explores Cosine and Euclidean similarity to adaptively explore the distance and angle relationship between pixels of different scale convolution patch features. Moreover, the Cross-Layer Information Complement module is designed to form a contextual interaction by integrating the output features of different convolution layers, which can prevent the omission of discriminative information and further improve the network performance. …”
Get full text

AnoViT: Unsupervised Anomaly Detection and Localization With Vision Transformer-Based Encoder-Decoder by Yunseung Lee, Pilsung Kang

“…Encoder-decoder structures have been widely used in the field of anomaly detection because they can easily learn normal patterns in an unsupervised learning environment and calculate a score to identify abnormalities through a reconstruction error indicating the difference between input and reconstructed images. Therefore, current image anomaly detection methods have commonly used convolutional encoder-decoders to extract normal information through the local features of images. …”
Get full text

Speech Recognition in an Enclosure with a Long Reverberation Time by Jedrzej KOCINSKI, Edward OZIMEK

“…Impulse Responses (IRs) were first determined with a dummy head in different measurement points of the enclosure. The following objective parameters were calculated with Dirac 4.1 software: Reverberation Time (RT), Early Decay Time (EDT), weighted Clarity (C$_{50}$) and Speech Transmission Index (STI). …”
Get full text

Adaptation of image reconstruction algorithm for purposes of ultrasound transmission tomography (UTT) by A. B. DOBRUCKI, K. J. OPIELIŃSKI

“…In particular, a complete computer algorithm enabling the use of different convolving and interpolation functions has been developed. …”
Get full text

ShipYOLO: An Enhanced Model for Ship Detection by Xu Han, Lining Zhao, Yue Ning, Jingfeng Hu

“…In the training process, the 3 × 3 convolution, 1 × 1 convolution, and identity parallel mode are used to replace the original feature extraction component (ResUnit) and more features are extracted. …”
Get full text

Deep Learning Model of Image Classification Using Machine Learning by Qing Lv, Suzhen Zhang, Yuechun Wang

“…Firstly, based on the analysis of the basic theory of neural network, this paper expounded the different types of convolution neural network and the basic process of its application in image classification. …”
Get full text

SFEF-Net: Scattering Feature Extraction and Fusion Network for Aircraft Detection in SAR Images by Qiang Zhou, Zongxu Pan, Ben Niu

“…Firstly, we proposed an innovative sparse convolution operator and applied it to feature extraction. …”
Get full text

Classification of Satellite Image Time Series and Aerial Images Based on Multiscale Fusion and Multilevel Supervision by H. Kanyamahanga, M. Dorozynski, F. Rottensteiner

“…In this context, it is a challenge to train a classifier given the large difference in resolutions. We utilise convolutions to extract spatial information and consider self-attention in the temporal dimension for SITS. …”
Get full text

On Symmetrical Sonin Kernels in Terms of Hypergeometric-Type Functions by Yuri Luchko

“…In this paper, a new class of kernels of integral transforms of the Laplace convolution type that we named symmetrical Sonin kernels is introduced and investigated. …”
Get full text

Scene Text Detection Based on Multi-scale Feature Extraction and Bidirectional Feature Fusion by LIAN Zhe, YIN Yanjun, ZHI Min, XU Qiaozhi

“…However, single-scale convolution methods are usually difficult to take into account the feature representation of text targets with different shapes and scales. …”
Get full text

Two-stage Detection Method for Abnormal Cluster Cervical Cells by LIANG Yi-qin, ZHAO Si-qi, WANG Hai-tao, HE Yong-jun

“…The size and location of convolution kernel can be dynamically adjusted according to the current pathological image content, so as to adapt to the shape, size and other geometric changes of cervical cells in different clusters. …”
Get full text

Reversible image steganography based on residual structure and attention mechanism by Lianshan Liu, Shanshan Tong, Qianwen Xue

“…The network uses INN as the overall framework, adopts a double-branch structure, extracts deep features using the mixed attention mechanism, and employs channel shuffle to promote information interaction between different features. This paper introduces dilated convolution to design a multi-scale convolution attention module that combines feature information from different scales, highlights essential features, and precisely locates the ideal embedding position. …”
Get full text

The TDGL Module: A Fast Multi-Scale Vision Sensor Based on a Transformation Dilated Grouped Layer by Leilei Xie, Fenghua Zhu, Zhixue Wang

“…These improvements enable the network to distinguish features at different scales effectively while optimizing spatial information processing and reducing computational costs. …”
Get full text

Reduction of Electromagnetic Reflections in 3D Airborne Transient Electromagnetic Modeling: Application of the CFS-PML in Source-Free Media by Yanju Ji, Xuejiao Zhao, Jiayue Gu, Dongsheng Li, Shanshan Guan

“…To solve the problem of electromagnetic reflections caused by the termination of finite-difference time-domain (FDTD) grids, we apply the complex frequency-shifted perfectly matched layer (CFS-PML) to airborne transient electromagnetic (ATEM) modeling in a source-free medium. …”
Get full text

YOLO-UIR: A Lightweight and Accurate Infrared Object Detection Network Using UAV Platforms by Chao Wang, Rongdi Wang, Ziwei Wu, Zetao Bian, Tao Huang

“…Moreover, the LSP module efficiently combines features from different distances using Large Receptive Field Convolution Layers, significantly enhancing the model’s long-range information capture capability. …”
Get full text

HRNet Encoder and Dual-Branch Decoder Framework-Based Scene Text Recognition Model by Meiling Li, Xiumei Li, Junmei Sun, Yujin Dong

“…In the decoder module, the dual-branch structure is adopted, in which the super-resolution branch takes the feature maps with the highest resolution obtained in the encoder module as input and restores images by upsampling through transposed convolution. The four kinds of feature maps with different resolutions are fused through independent transposed convolution layers for multiscale fusion in the recognition branch and then inputted into the attention-based decoder for text recognition. …”
Get full text

Learner preferences prediction with mixture embedding of knowledge and behavior graph by Xiaoguang LI, Lei GONG, Xiaoli LI, Xin ZHANG, Ge YU

“…To solve the problems of inaccurate prediction of learner preference and insufficient utilization of structural information in the knowledge recommendation model, for the knowledge structure and learner behavior structure in the learner’s preference prediction model, the model of learner preferences predication with mixture embedding of knowledge and behavior graph was proposed.First, considering using graph convolution network (GCN) to fit structural information, GCN was extended to knowledge graph and behavior graph, the purpose of which was to obtain learners’ overall learning pattern and individual learning pattern.Then, the difference between knowledge structure and behavior structure was used to fit learners’ individual preferences, and recurrent neural network was used to encode and decode learners’ preferences to obtain the distribution of learners’ preference distribution.The experimental results on the real datasets demonstrate that the proposed model has a good effect on predicting learner preferences.…”
Get full text

Research on the Classification Method of Tea Tree Seeds Quality Based on Mid-Infrared Spectroscopy and Improved DenseNet by Di Deng, Hao Li, Jiawei Luo, Jiachen Jiang, Hongbo Mu

“…Four types of tea tree seeds in different states were prepared, and their spectral data were collected and preprocessed using Savitzky–Golay (SG) filtering and wavelet transform. …”
Get full text