A5054991988- Nasal Surgery and Airway Studies
- Image Processing and 3D Reconstruction
- Industrial Engineering and Technologies
- Advanced Sensor and Control Systems
- Underwater Acoustics Research
- Remote Sensing and Land Use
- Advanced SAR Imaging Techniques
- Obstructive Sleep Apnea Research
- Educational Technology and Assessment
- Sparse and Compressive Sensing Techniques
- Arctic and Antarctic ice dynamics
- Image and Object Detection Techniques
- Handwritten Text Recognition Techniques
- Advanced Computational Techniques in Science and Engineering
- Synthetic Aperture Radar (SAR) Applications and Techniques
- Sinusitis and nasal conditions
- Advanced Algorithms and Applications
- Advanced Data Processing Techniques
- Environmental Sustainability and Technology
Beijing Technology and Business University
2023-2024
Hangzhou Normal University
2023
In recent years, the advancement of deep learning technology has led to excellent performance in synthetic aperture radar (SAR) automatic target recognition (ATR) technology. However, due interference speckle noise, task classifying SAR images remains challenging. To address this issue, a multi-scale local–global feature fusion network (MFN) integrating convolution neural (CNN) and transformer was proposed study. The comprises three branches: CovNeXt-SimAM branch, Swin Transformer branch....
The paper proposes a deep neural network architecture based on the integration of convolutional Faster R-CNN with Feature Pyramid Network module. Based this approach, an algorithm for detecting and classifying vehicles in images corresponding model have been developed.
 A cross-platform environment ML.NET was used to train proposed model. results comparing effectiveness approach networks YOLO v4 are presented. improvement accuracy detection localization different types ultra-high...
<title>Abstract</title> To improve the accuracy of offline handwritten Chinese character recognition (offline HCCR), combined with self-attention, this paper proposes a collaborative multi-model approach for HCCR. Most existing HCCR models use different network structures to obtain features, which may lead results in terms accuracy, and that errors occurred also be different. Utilizing feature, we propose our method hoping their accuracy. In paper, five models, including AlexNet, VGG16,...
The nasal cavity three-dimensional structure reconstruction is significant for the diagnosis and treatment of nasopharynx diseases. Currently, most systems are based on a long-source acoustic tube. However, tube has weaknesses low portability high cost. To address these issues, system short-source was designed in this research. comprises lower computer signal acquisition device upper analysis software. consists processing unit, detection microcontroller, USB communication interface. It used...