检索结果-内蒙古大学图书馆

deep rigid registration for slice-to-volume in real time

EXPERT SYSTEMS WITH APPLICATIONS 2024年 235卷

作者： Zheng, Shiqiang Wang, Yifan Yang, Xin Deng, Xianbo Ding, Mingyue Hou, Wenguang Huazhong Univ Sci & Technol Coll Life Sci & Technol Wuhan 430074 Hubei Peoples R China Sun Yat Sen Univ Collaborat Innovat Ctr Canc Med State Key Lab Oncol South China Canc CtrGuangdong Key Lab Nasopharyngeal Carcinom Guangzhou 510060 Guangdong Peoples R China Huazhong Univ Sci & Technol Union Hosp Radiol Dept Wuhan Hubei Peoples R China

Slice-to-volume registration that achieves high accuracy in real time is one of the enabling technologies for clinical imaging scenarios. Recently, deep learning methods have been explored to significantly improve the accuracy and efficiency of the registration. Nevertheless, such a 2D/3D registration problem is very challenging due to several considerable barriers including highly computational cost brought by dense sampling in six dimensional parameter space and significant modal appearance differences. We proposed a Differentiable resampling based Slice-to-Volume Registration network, which can achieve real-time and accurate registration in both mono-and multi-modal scenarios. The proposed network learns the out-of-plane transformation parameters in the spherical coordinates that decide the content and is invariant to the remaining in-plane parameters by feeding 2D rigid transformed data on-the-fly, thus reducing the training sample size remarkably. To further improve the performance, we introduce an auxiliary image similarity connected by differentiable resampling. For multi-modal scenarios, we optionally include the enhanced modality independent neighborhood descriptor to unify different modalities into common space. Training size is considerably reduced to 30k per dataset with half of all possible orientations and 80% of the volume space covered while the average registration error drops to 1.84 mm and 6.49 mm respectively for mono-and multi-modal experiments. Extensive experiments show the proposed methods' superiority over existing deep learning methods for real-time 2D/3D rigid registration in both mono-and multi-modal settings.

关键词： 2-D/3-D registration Convolutional neural network deep learning image guided intervention

来源：评论

学校读者我要写书评

暂无评论

Detecting particle quantity from raw reconstructed images using digital holography and deep learning

引用

OPTICAL ENGINEERING 2024年第7期63卷 73102-73102页

作者： Li, Wei-Na Ou, Hongjie Liao, Junpeng Xie, Xiangsheng Shantou Univ Coll Sci Phys Dept Shantou Peoples R China

We propose a modified residual neural network (ResNet) to quickly and accurately detect the particle quantity from raw reconstructed images of a high-density particle solution in digital holography. The raw reconstructed image is fed into the modified ResNet to obtain the particle quantity. Then, the quantity and particle concentration in the captured volume are calculated. The metrological challenge is modeled as a regression problem in deep learning. The average relative error of the holograms in the test dataset is less than 10% even when predicting the particle quantities untrained by the model. Hence, an accurate particle quantity is obtained even when the raw reconstructed images are not denoised, thereby reducing the processing time.

关键词： digital holography Fresnel diffraction deep learning residual neural network quantity measurement particles

来源：评论

学校读者我要写书评

暂无评论

deep learning-based method for real-time spinach seedling health monitoring

引用

JOURNAL OF ELECTRONIC IMAGING 2024年第2期33卷

作者： Xu, Yanlei Cong, Xue Zhai, Yuting Bai, YuKun Yang, Shuo Li, Jian Jilin Agr Univ Dept Informat & Technol Changchun Peoples R China

It is difficult to obtain real-time crop water shortage during spinach seedling production, so this study proposes a real-time spinach seedling water stress classification method based on hierarchical transformer encoder (HTE) MobileNet (HT-MobileNet). An efficient sandglass residual module is designed for the grading algorithm to significantly improve the model grading accuracy. Meanwhile, a lightweight attention mechanism is proposed. It can effectively suppress the influence of irrelevant features on the classification results. Furthermore, an HTE is introduced, classifying the degree of stress accurately under extremely complex spinach seedling conditions. By introducing grouped convolution, the number of model parameters are reduced greatly, which lays the foundation for practical application. The grading accuracy is 94.61%. The number of parameters is 1.78M and the model size is 4.57MB, which are 40.67% lower than the original network. Comparison experiments are conducted between HT-MobileNet and 10 mainstream convolutional neural networks, such as ConvNext. The experimental results demonstrate the outstanding advantages of this research in terms of accuracy and weight. In addition, an intelligent spinach seedling water stress grading system is designed and tested. This study can provide technical support for real-time monitoring of spinach seedling growth status. (c) 2024 SPIE and IS&T

关键词： convolutional neural network water stress classification image processing complex backgrounds

来源：评论

学校读者我要写书评

暂无评论

real-time REMOTE MONITORING AND TRACKING TECHNOLOGY FOR THREE-DIMENSIONAL WAREHOUSE BASED ON deep learning AND TARGET DETECTION

UPB Scientific Bulletin, Series C: Electrical Engineering an...

引用

UPB Scientific Bulletin, Series C: Electrical Engineering and Computer Science 2024年第2期86卷 211-230页

作者： Yan, Shaokui Ding, Haili Zhang, Jie Wang, Xiangwei Wang, Mingqiang Yinchuan China Comarvel Intelligent Technology Company Limited Yantai China

In order to realize the dynamic management of goods in the warehouse and improve the overall management efficiency and safety performance of the warehouse, a deep learning and target detection technology for remote real-time monitoring and tracking of the stereo warehouse is proposed. Wireless sensor nodes are used to provide all-round coverage of the three-dimensional warehouse to meet the monitoring needs of various parameters such as cargo status and cargo location, and the image data of the three-dimensional warehouse is trained by deep confidence networks to extract high-quality and distinguished features of the three-dimensional warehouse. The hybrid Gaussian model is used to accurately locate and identify the target warehouse, determine the location and trajectory of warehouse objects, and finally introduce a priori and measurement information by virtue of Kalman filtering method to realize real-time monitoring and tracking of the stereo warehouse. The results show that the AUC value of the complex monitoring attributes of the proposed method is as high as 0.93 after the optimization process, and the monitoring accuracy is as high as 98.43%, and the monitoring time is short, and the predicted value of the center of mass coordinates of the warehouse is the same as the real value of the actual coordinates, which indicates that the proposed method can ensure the comprehensiveness and real-time monitoring and tracking, and realize the intelligence and efficiency of the warehouse management. © 2024, Politechnica University of Bucharest. All rights reserved.

关键词： Target tracking

来源：评论

学校读者我要写书评

暂无评论

Smarter Aging: Developing a Foundational Elderly Activity Monitoring System With AI and GUI Interface

引用

IEEE ACCESS 2024年 12卷 74499-74523页

作者： Htet, Ye Zin, Thi Thi Tin, Pyke Tamura, Hiroki Kazuhiro, Kondo Watanabe, Shinji Chosa, Etsuo Univ Miyazaki Interdisciplinary Grad Sch Agr & Engn Miyazaki 8892192 Japan Univ Miyazaki Grad Sch Engn Miyazaki 8892192 Japan Univ Miyazaki Fac Med Miyazaki 8892192 Japan

The global rise in the elderly population, which presents challenges to healthcare systems owing to labor shortages in caregiving facilities, necessitates innovative solutions for elderly care services. Smart aging technologies such as robotic companions and digital home gadgets, offer a solution to these challenges by improving the elderly's quality of life and assisting caregivers. However, limitations in data privacy, real-time processing, and reliability often hinder the effectiveness of the existing technologies. Among these, privacy concerns are a major barrier to ensuring user trust and ethical implementation. Therefore, this study proposes a more effective approach for smart aging through elderly activity monitoring that prioritizes data privacy. The proposed system utilizes stereo depth cameras to monitor the activities of the elderly. Data were collected from real-world environments with the participation of six elderly individuals from a care center and hospital. This system focuses on recognizing common daily actions of the elderly including sitting, standing, lying down, and seated in a wheelchair. Additionally, it recognizes transition states (in-between actions such as changing from sitting to standing) that are crucial for assessing balance issues. By integrating motion information with a deep-learning architecture, the system achieved a high accuracy of 99.42% in recognizing daily actions in real-time. This high accuracy was maintained even with minimal data from new environments through transfer learning, and the adaptability of this model ensured its potential for real-world applications. For intuitive interaction between the caregivers and the system, a user-friendly graphical interface (GUI) was also designed in the proposed approach.

关键词： Older adults Monitoring Aging Cameras Medical services Feature extraction Ethics deep learning Motion control real-time systems deep learning architecture elderly activity monitoring GUI motion information real-time action recognition smart aging stereo depth cameras transition state recognition

来源：评论

学校读者我要写书评

暂无评论

real-time Myocardial Infarction Detection Approaches with a Microcontroller-Based Edge-AI Device

引用

SENSORS 2024年第3期24卷 828-828页

作者： Gragnaniello, Maria Borghese, Alessandro Marrazzo, Vincenzo Romano Maresca, Luca Breglio, Giovanni Irace, Andrea Riccio, Michele Univ Naples Federico II Dept Elect Engn & Informat Technol DIETI I-80125 Naples Italy

Myocardial Infarction (MI), commonly known as heart attack, is a cardiac condition characterized by damage to a portion of the heart, specifically the myocardium, due to the disruption of blood flow. Given its recurring and often asymptomatic nature, there is the need for continuous monitoring using wearable devices. This paper proposes a single-microcontroller-based system designed for the automatic detection of MI based on the Edge Computing paradigm. Two solutions for MI detection are evaluated, based on Machine learning (ML) and deep learning (DL) techniques. The developed algorithms are based on two different approaches currently available in the literature, and they are optimized for deployment on low-resource hardware. A feasibility assessment of their implementation on a single 32-bit microcontroller with an ARM Cortex-M4 core was examined, and a comparison in terms of accuracy, inference time, and memory usage was detailed. For ML techniques, significant data processing for feature extraction, coupled with a simpler Neural Network (NN) is involved. On the other hand, the second method, based on DL, employs a Spectrogram Analysis for feature extraction and a Convolutional Neural Network (CNN) with a longer inference time and higher memory utilization. Both methods employ the same low power hardware reaching an accuracy of 89.40% and 94.76%, respectively. The final prototype is an energy-efficient system capable of real-time detection of MI without the need to connect to remote servers or the cloud. All processing is performed at the edge, enabling NN inference on the same microcontroller.

关键词： deep learning edge computing machine learning myocardial infarction detection

来源：评论

学校读者我要写书评

暂无评论

A multi-view ensemble model based on semi-supervised feature learning for small sample classification of PolSAR images

引用

INTERNATIONAL JOURNAL OF REMOTE SENSING 2024年第3期45卷 981-1031页

作者： Darvishnezhad, Mohsen KN Toosi Univ Technol Dept Elect Engn Tehran Iran

Classification with a few samples of training set has been a long-standing issue in the field of polarimetric synthetic aperture radar (PolSAR) image analysis and processing. In fact, one of the most important challenges of the PolSAR image classification task is the number of labelled samples. In essence, training network by utilizing just few numbers of training samples cannot be led to the reliable result since the neural network is so sensitive to the number of training samples. In addition, in the real PolSAR image classification task, the huge number of training samples is not accessible. On the other hand, classification performance using just small number of labelled samples is not an accurate result. So, in this paper, aiming at the small number of training samples of the PolSAR image classification task, a novel ensemble self-supervised feature-learning (ESSFL) model is designed. The designed ESSFL can automatically extract PolSAR features conducive to PolSAR image classification with a small number of training samples. In addition, it can significantly decrease the dependence of neural network algorithms on large labelled samples of training set. First, to utilize the spatial-polarimetric features of PolSAR data perfectly, the EfficientNet-B0 is presented and utilized as the main section of the deep learning (DL) model to extract DL features of PolSAR data. Then, using an optimization function that constrains the cross-correlation matrix of various distortions of each sample to the identity matrix, the designed deep learning model can obtain the effective features of homogeneous samples gathering together and heterogeneous samples separating from each other in a self-supervised manner. Moreover, two ensemble learning (EL) models, feature-level and view-level ensemble, are proposed to increase the feature extraction capability and classification result by jointly using spatial features at different scales and polarimetric information at different bands. Fi

关键词： Polarimetric synthetic aperture radar image classification small sample classification self-supervised learning ensemble learning deep learning

来源：评论

学校读者我要写书评

暂无评论

Wireless channel corrupted image denoising using residual learning of adaptive wavelet with dilated deep convolutional neural network

引用

SIGNAL image AND VIDEO processing 2024年第3期18卷 2309-2321页

作者： Sreedhar, M. Aparna, K. Jawaharlal Nehru Technol Univ Anantapur Ananthapuramu Andhra Pradesh India Jawaharlal Nehru Technol Univ Anantapur JNTUA Coll Engn ECE Dept Constituent Coll Kalikiri Ananthapuramu Andhra Pradesh India

In digital image transmission, channel coding is essential for data integrity and accurate reception during digital communication. While minimizing the consequences of the channel, typical communication methods neglect the received context and content-related data. Many significant restrictions are applied to image transmission in digital communication. Due to the limited qualities of transferred data, image quality suffers across over wireless networks. Moreover, the transmission of digital images is a complicated task based on their shape, size, and bandwidth. Hence, it lacks flexibility and practicality in the real-world environment. Different image-denoising techniques are employed to decrease the noisy image channel effects. Thus, this research aims to analyze the images gathered during wireless channel communication to precisely eliminate the bugs and the impact of channel degradation. The image denoising is carried out during the wireless channel communication process at the receiver end. A novel deep Residual learning of Adaptive Wavelet with Dilated deep Convolutional Neural Network (RL-AWDDCNN) method is used to denoise the image more effectively. Hence, the residual images are generated based on the heuristic concept, where an improved flow direction algorithm (IFDA) is developed for optimizing the wavelet parameters with Dilated deep Convolutional Neural Network. Throughout the result analysis, the designed method scored a 3037.233% peak signal-to-noise ratio (PSNR) rate and an 80.18675% structural similarity index measure (SSIM) rate. Thus, the developed denoising model's performance and the experimental results are analyzed and compared with various existing methods concerning the standard image quality measures.

关键词： image denoising Dilated deep convolutional neural network Improved flow direction algorithm Discrete wavelet transform image decomposition

来源：评论

学校读者我要写书评

暂无评论

learning global and local features of power load series through transformer and 2D-CNN: An image-based multi-step forecasting approach incorporating phase space reconstruction

引用

APPLIED ENERGY 2024年 378卷

作者： Tang, Zihan Ji, Tianyao Kang, Jiaxi Huang, Yunlin Tang, Wenhu South China Univ Technol Sch Elect Power Engn Guangzhou 510640 Peoples R China

As modern power systems continue to evolve, accurate power load forecasting remains a critical issue in energy management. The phase space reconstruction (PSR) method can effectively retain the inner chaotic property of power load from a system dynamics perspective and thus is a promising knowledge-based method for power load forecasting. To fully leverage the PSR method's capability in modeling this high-dimensional, non-stationary characteristics of power load data, and to address the challenges faced by its classical mathematical prediction algorithms ineffectively solving contemporary prediction scenarios characterized by massive volumes of data. This study proposes a novel learning-based multi-step forecasting approach that utilizes an image-based modeling perspective for the reconstructed phase trajectories. Firstly, the feature engineering approach that simultaneously utilizes dynamic evolution features and temporal locality features in the trajectory image is proposed. Through mathematical derivation, the equivalent characterization of the PSR method and another time series modeling approach, patch segmentation (PS), is demonstrated for the first time. Building on this prior knowledge, a novel image-based modeling perspective incorporating a global and local feature extraction strategy is introduced to fully leverage these valuable features. Subsequently, within this framework, a novel deep learning model, termed PSR-GALIEN, is designed for end-to-end processing. This model employs a Transformer Encoder and 2D convolutional neural networks (CNNs) to extract global and local patterns from the image, while a multi-layer perceptron (MLP)-based predictor is utilized for efficient correlation modeling. Extensive experiments on five real-world datasets show that PSR-GALIEN consistently outperforms six state-of-the-art deep learning models in short-term load forecasting scenarios with varying characteristics, demonstrating its great robustness. Ablation studies fur

关键词： Multi-step power load forecasting Phase space reconstruction image-based modeling perspective Global and local feature extraction Feature interpretation

来源：评论

学校读者我要写书评

暂无评论

real time Pothole Detection System Using deep learning Techniques: A Systematic Review 4

Real Time Pothole Detection System Using Deep Learning Techn...

引用

4th International Conference on Ubiquitous Computing and Intelligent Information Systems, ICUIS 2024

作者： Saraswathi, S. Vigneshwari, S. Sathyabama Institute of Science and Technology Department of Computer Science and Engineering Tamil Nadu Chennai India

ISBN: (纸本)9798331529635

real-time pothole detection serves a crucial role in maintaining road infrastructure and ensuring the safety of drivers. Traditional methods for identifying potholes are often labour-intensive and inefficient, prompting the exploration of automated techniques using deep learning. This comprehensive review aims to synthesize the current state of research on pothole detection systems leveraging deep learning algorithms. By examining a range of studies published over the past decade, the review highlights various methodologies, including You Only Look Once (YOLO), convolutional neural networks (CNNs), deep CNN, and hybrid models, as well as the different datasets and image processing techniques employed. The review categorizes the approaches based on their accuracy, computational efficiency, and practical implementation challenges. The review also identifies research gaps in the literature survey, such as the demand for standardized datasets and real-time processing capabilities, and suggests directions for future research. By providing a systematic overview of deep learning-based pothole detection systems, the purpose of this review is to help researchers and practitioners develop more reliable and scalable solutions for maintaining road infrastructure. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：