检索结果-内蒙古大学图书馆

Conference on Polarized Light and Optical Angular Momentum for Biomedical Diagnostics

作者： Si, Lu Huang, Tongyu Wang, Xingjian Yao, Yue Ma, Hui Tsinghua Univ Tsinghua Berkeley Shenzhen Inst Shenzhen 518071 Peoples R China Tsinghua Univ Shenzhen Int Grad Sch Shenzhen 518055 Peoples R China Tsinghua Univ Dept Biomed Engn Beijing 100084 Peoples R China Tsinghua Univ Dept Phys Beijing 100084 Peoples R China

ISBN: (纸本)9781510647985;9781510647978

Polarization is capable of probing microstructures and has unique sensitivity to fibrous anisotropic structure. Polarimetric imaging has demonstrated promising potential in diverse applications ranging from biomedicine, material science, and atmospheric remote sensing. The polarization properties of samples can be comprehensively described by a Mueller matrix (MM). However, the relationship between individual MM elements and properties of the sample is often not clear. There have been consistent efforts to derive polarization parameters from MM based on certain assumptions for better description of the samples, e.g., MM polar decomposition (MMPD), MM transformation (MMT) and MM differential decomposition. Usually, the MM imaging requires sequential measurements with different polarization states of incident light and the imaging process is time consuming. In addition, for movable samples, we cannot guarantee the consistency during the imaging. This may cause precision issues since the images cannot be well-registered. In this work, we built a statistical translation model to generate polarization parameters from a single Stokes vector which can be obtained by one-shot imaging. This will improve the imaging efficiency, simplify the optical system and avoid introducing errors by the image registration. In the model design, we adopted the generative adversarial network (GAN) where the generator is based on a U-net architecture. We demonstrated the effectiveness of our approach on liver tissue, blood smear and porous anodic alumina (PAA) film, and quantitatively evaluated the results by similarity assessment methods. The model can generate a parameter image within 0.1 second on a desktop computer, which shows the potential to achieve real-time performance.

关键词： Polarimetric imaging deep learning Mueller matrix Stokes vector

来源：评论

学校读者我要写书评

暂无评论

SAR Scene Classification Based on Self-supervised Jigsaw Puzzles 5th

SAR Scene Classification Based on Self-supervised Jigsaw Puz...

引用

5th IFIP TC 12 International Conference on Intelligence Science, ICIS 2022

作者： Ren, Zhongle Lu, Yiming Wang, Hanxiao Zhang, Yu Hou, Biao Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education of China Xidian University Xi’an710071 China

ISBN: (纸本)9783031149023

Scene classification is a hot issue in the field of SAR image interpretation. Many SAR image interpretation tasks can be promoted with the development of highly credible scene classification methods. But the fussy steps of traditional methods and the imperious demands of labeled samples in deep learning-based methods restrict the effective feature learning in SAR scene classification. Hence, a self-supervised learning method based on Jigsaw puzzles is proposed to address the problems. Concretely, the Jigsaw puzzle reassembly of the SAR image block is firstly taken as the upstream task without manual labels. Once the correct spatial arrangement is obtained from it, the learned high-level feature from the upstream task is used as the pre-training model for the downstream task, which is then fine-tuned with only a few labeled samples to enhance the performance of the SAR scene classification task. Experimental results on 25-class real SAR scenes confirm the proposed method can greatly improve the scene classification performance than directly training the network with the same number of labeled samples used in the downstream task. © 2022, IFIP International Federation for Information processing.

关键词： Synthetic aperture radar

来源：评论

学校读者我要写书评

暂无评论

L-RTDETR: A Lightweight real-time Object Detection Algorithm for Defect Detection

L-RTDETR: A Lightweight Real-Time Object Detection Algorithm...

引用

International Conference on image, Vision and Computing (ICIVC)

作者： Xinran Chen Shufeng Li Wei Yang School of Information and Communication Engineering Communication University of China Beijing China School of Electronic Information Engineering Beihang University Beijing China

ISBN: (数字)9798350385991

ISBN: (纸本)9798350386004

In the current deep learning research landscape, image processing technology faces significant challenges in lightweight defect detection despite its broad application in object detection. This study introduces the Lightweight real-time DEtection TRansformer (L-RTDETR) model, achieving a balance between high accuracy and model lightness. The model employs a Large Selective Kernel Network (LSKNet) based lightweight backbone, reducing network size and parameters. It also integrates Deformable Attention (DAttention) based Intra-scale Feature Interaction to enhance detection speed with a sparse attention mechanism. By combining L1, Generalized Intersection over Union (GIoU), and Normalized Wasserstein Distance (NWD) loss functions with weighted metrics, the model’s adaptability for object detection tasks improves. Experimental results on the Northeastern University Defect Detection dataset show that L-RTDETR achieves a $35.8 \%$ reduction in model size and increases detection speed to 61.6 fps, outperforming the conventional RT-DETR algorithm. It also shows a $2.5 \%$ improvement in mean Average Precision at $0.5(m A P \text{@} 0.5)$ and a $1.7 \%$ increase in mAP@0.5-0.95. These results underscore the model’s efficiency in lightweight defect detection.

关键词： Measurement Adaptation models image synthesis Object detection Transformers Generative adversarial networks Feature extraction real-time systems Kernel Defect detection

来源：评论

学校读者我要写书评

暂无评论

Optimizing Gesture Recognition for Seamless UI Interaction Using Convolutional Neural Networks

Optimizing Gesture Recognition for Seamless UI Interaction U...

引用

image processing, Computer Vision and Machine learning (ICICML), International Conference on

作者： Qi Sun Tong Zhang Shang Gao Liuqingqing Yang Fenghua Shao Carnegie Mellon University Pittsburgh USA Loughborough University Loughborough United Kingdom Trine University Reston USA University of Michigan Ann Arbor Ann Arbor USA Independent Researcher Toronto Canada

ISBN: (数字)9798350355413

ISBN: (纸本)9798350355420

This study presents an advanced gesture recognition and user interface (UI) interaction system developed using deep learning technologies, emphasizing its transformative impact on UI design and functionality. By employing optimized convolutional neural networks (CNNs), the system achieves high-precision gesture recognition, significantly enhancing user interactions with digital interfaces. Initial steps involve preprocessing collected gesture images to conform to CNN input standards, followed by employing sophisticated feature extraction and classification methodologies. We address class imbalance effectively using Focal Loss as the loss function, ensuring robust model performance across varied gesture types. The experimental results showcase notable improvements in model metrics, with the AUC and Recall increasing progressively as we evolve from simpler models like VGG16 to more complex ones such as DenseNet. Our enhanced model demonstrates a significant advancement with an AUC of 0.83 and a Recall of 0.85, surpassing standard benchmarks. More critically, this system's capacity to support real-time and efficient gesture recognition paves the way for a new era in UI design—where intuitive, natural user gestures can seamlessly integrate into everyday technology use, significantly reducing the learning curve and enhancing user satisfaction. The broad implications of this development are profound, extending beyond mere technical performance to fundamentally reshape how users interact with technology. Such advancements hold considerable promise for the enhancement of smart life experiences, highlighting the pivotal role of gesture-based interactions in the next generation of UI development.

关键词： deep learning Measurement Human computer interaction Technological innovation Accuracy Refining Gesture recognition Feature extraction Convolutional neural networks Standards

来源：评论

学校读者我要写书评

暂无评论

Underwater image processing with New Dark Channel Prior Dehazing

Underwater Image Processing with New Dark Channel Prior Deha...

引用

International Conference on Communications, Information System and Computer Engineering (CISCE)

作者： Ruikang Hu Yuhan Li School of Information and Communication Guilin University of Electronic Technology Guilin China School of Mechanical and Electrical Engineering Guilin University of Electronic Technology Guilin China

Aiming at addressing the problems of low visibility and poor contrast, this paper proposes a new dark channel prior dehazing. According to the characteristics of the light source, an image is divided into light and non-light source areas. Mixed precision operation is used to subsample the dark channel image and deep learning network and GPU-accelerated method are used to improve the algorithm speed to solve the real-time problem. Experimental results show that compared with similar algorithms, the new algorithm is more balanced in image quality indicators and underwater image indicators, which better working requirements of underwater vehicles. In terms of real-time performance, the new algorithm is superior to similar algorithms. When processing images a $950\times 550$ pixel, resolving new with an average frame rate of 29.4, which runs 2.46 times faster than dark channel prior, which lays a foundation for underwater robots to carry out underwater operations more efficiently.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Autonomously Improving Systems in Industry: A Systematic Literature Review 1

引用

11th International Conference on Software Business

作者： Green, Rolf Bosch, Jan Olsson, Helena Holmstrom Chalmers Univ Technol Gothenburg Sweden Malmo Univ Malmo Sweden

ISBN: (数字)9783030672928

ISBN: (纸本)9783030672911;9783030672928

A significant amount of research effort is put into studying machine learning (ML) and deep learning (DL) technologies. real-world ML applications help companies to improve products and automate tasks such as classification, image recognition and automation. However, a traditional "fixed" approach where the system is frozen before deployment leads to a sub-optimal system performance. Systems autonomously experimenting with and improving their own behavior and performance could improve business outcomes but we need to know how this could actually work in practice. While there is some research on autonomously improving systems, the focus on the concepts and theoretical algorithms. However, less research is focused on empirical industry validation of the proposed theory. Empirical validations are usually done through simulations or by using synthetic or manually alteration of datasets. The contribution of this paper is twofold. First, we conduct a systematic literature review in which we focus on papers describing industrial deployments of autonomously improving systems and their real-world applications. Secondly, we identify open research questions and derive a model that classifies the level of autonomy based on our findings in the literature review.

关键词： Autonomously improving systems Machine learning Industrial application AI engineering Empirical validation

来源：评论

学校读者我要写书评

暂无评论

Research on Water Conservancy Project Monitoring and Intelligent Identification Method Based on Computer Vision Technology

Research on Water Conservancy Project Monitoring and Intelli...

引用

Telecommunications and Power Electronics (TELEPE), International Conference on

作者： Jin Zhu Feng Yang Zhejiang Province Water Conservancy and Hydropower Technology Consultation Center Hangzhou Zhejiang China Zhejiang Zhongshui Engineering Technology Co. Ltd. Hangzhou China

ISBN: (数字)9798350369212

ISBN: (纸本)9798350380002

Monitoring and maintenance of water resources projects is essential to ensure project safety and environmental protection. Traditional monitoring methods often rely on manual inspections and sensor data, but these methods suffer from high cost, low efficiency and limited monitoring range. With the rapid development of computer vision technology, its potential for application in water conservancy engineering has gradually emerged. The purpose of this paper is to study the water conservancy project monitoring and intelligent identification methods based on computer vision technology. We first outline the basic concepts of computer vision and its application scenarios in water conservancy engineering, including structural health monitoring, environmental monitoring and disaster management. Then, the specific applications of image processing techniques, machine learning and deep learning in intelligent recognition are described in detail, especially how these techniques can be utilized to achieve an efficient and accurate real-time monitoring system. Through several case studies, this paper demonstrates the effects and advantages of the application of computer vision technology in actual water conservancy projects. Finally, it summarizes the main findings of the current study and looks forward to the future technological development and application prospects, pointing out the direction of further research. The findings of this paper show that computer vision-based monitoring and intelligent identification methods not only improve the accuracy and efficiency of monitoring, but also provide new technical means for the management and maintenance of water conservancy projects.

关键词： deep learning Computer vision Accuracy Water conservation real-time systems Maintenance Floods Protection Monitoring Water resources

来源：评论

学校读者我要写书评

暂无评论

Review on the application of deep learning in network attack detection

引用

JOURNAL OF NETWORK AND COMPUTER APPLICATIONS 2023年 212卷 1-15页

作者： Yi, Tao Chen, Xingshu Zhu, Yi Ge, Weijing Han, Zhenhui Sichuan Univ Sch Cyber Sci & Engn Chengdu Peoples R China China Elect Technol Cyber Secur Co Ltd Ind Dev Dept Chengdu Sichuan Peoples R China Sichuan Univ First Ring Rd24 South Sect Chengdu Sichuan Peoples R China

With the development of new technologies such as big data, cloud computing, and the Internet of Things, network attack technology is constantly evolving and upgrading, and network attack detection technology is forced to undergo corresponding iterative evolution. Three main problems are associated with these technologies: the automatic representation of heterogeneous and complex network traffic data, the uneven network attack samples, and the contradiction between the accuracy of the anomaly detection model and the continuous evolution of attacks. Researchers have proposed several network attack detection techniques based on deep learning to address these problems. This study reviews and analyzes the studies aimed at dealing with such problems, considering multiple factors, such as models, traffic representation and feature extraction, threat detection model training, and model robustness improvement. Finally, the existing problems and challenges associated with the current research are analyzed with respect to data category imbalance, high-dimensional massive data processing, concept distribution drift, real-time interpretability of the detection model, and the security of the model.

关键词： Network attack deep learning Flow characterization Detection model Model security

来源：评论

学校读者我要写书评

暂无评论

深度学习语义分割方法用于双致密天体并合引力波数据处理研究

引用

天文学进展 2024年第2期42卷 299-314页

作者：马存良钟国健闵源嘉明珍贺观圣江西理工大学赣州341000 南华大学衡阳421001 中国科学院紫金山天文台南京210023

研究表明,基于深度学习的引力波搜寻,有望解决匹配滤波方法的低效率问题。与传统的匹配滤波引力波信号搜寻、基于贝叶斯后验或深度学习的波源参数估计方法相比,基于深度学习的端对端的引力波搜寻输出缺乏波源的时频信息,如双星并合信号... 详细信息

研究表明,基于深度学习的引力波搜寻,有望解决匹配滤波方法的低效率问题。与传统的匹配滤波引力波信号搜寻、基于贝叶斯后验或深度学习的波源参数估计方法相比,基于深度学习的端对端的引力波搜寻输出缺乏波源的时频信息,如双星并合信号到达时间和信号在探测站灵敏度区间的持续时间等。首次实现了深度学习用于应变信号时频点的细粒度分类,研究了基于图像语义分割的引力波搜寻方法。通过真实噪声和仿真物理信号合成数据,使用Q变换将其转换为时频图,构建像素级标注数据集。构建并训练时频图像语义分割模型,研究了其应用于引力波信号数据分析的可行性。结果表明语义分割方法未来有望作为基于深度学习的引力波信号搜寻流水线的一个组件,还可用于进一步提取信号的时频信息。

关键词：引力波数据分析深度学习语义分割

来源：评论

学校读者我要写书评

暂无评论

Over-the-Air Radar Emitter Signal Classification Based on SDR 6

Over-the-Air Radar Emitter Signal Classification Based on SD...

引用

6th IEEE International Conference on Intelligent Computing and Signal processing, ICSP 2021

作者： Xia, Yan Ma, Zhiyuan Huang, Zhi Naval University of Engineering School of Electronic Engineering Wuhan China

ISBN: (纸本)9780738143705

At present, in the field of radar emitter classification, theoretical simulation is mostly used to carry out algorithm research. However, there are few schemes to study signal classification in real electromagnetic environment using actual hardware. Therefore, this paper proposes a radar emitter classification scheme based on HackRF Software Defined Radio (SDR) and deep learning to solve the problem of weak engineering practice. Firstly, the GNU Radio development environment is used to realize the integration design of real space signal transceiver and time-frequency analysis algorithm application on HackRF hardware platform. Then, a classification model with 11 layers network is constructed to automatically extract the deep features of intra-pulse signal time-frequency image. Finally, the classification performance of eight kinds of signals in real electromagnetic environment is tested. The total recognition accuracy of this scheme is more than 83% under 6dB low Signal-to-Noise Ratio (SNR), which proves the effectiveness of the scheme, and provides an important basis for practical engineering application in the future. © 2021 IEEE.

关键词： deep learning

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：