检索结果-内蒙古大学图书馆

Detection of Group Test Objects Using Neural Networks with Hyperspectral Labeling During External Trajectory Measurements

引用

MEASUREMENT TECHNIQUES 2022年第3期65卷 166-173页

作者： Kuleshov, Ivan A. Minist Def Russian Federat Mil Unit 03080 Priozersk Kazakhstan

The task of detecting elements of group flight test objects is considered. The relevance of the research is related to the growing need for metrological support of external trajectory measurements when testing advanced multi-agent systems, which comprise aircrafts as elements. Existing algorithms for detecting objects used in modern optoelectronic stations (systems) are analyzed. According to the conducted analysis, the failure of these algorithms to meet the requirements imposed on testing group objects is largely due to the complexity of selecting objects against the atmospheric background. In order to ensure efficient detection of group objects, it is proposed to use additional highly-informative features that reflect the physical nature of test objects, such as the shape of the spectrum of light reflected from various materials. Such features can be identified in the process of preliminary hyperspectral analysis of test objects with a subsequent reduction in the data dimensionality. In order to eliminate errors arising during the rotation and partial overlapping of test objects, it is proposed to use neural network methods - in particular, the neural network detector YOLO v2. An approach to organizing data for training this detector is proposed along with an architecture that provides high accuracy and speed of information processing. These detector properties were confirmed experimentally on the basis of images exported from a model of external trajectory measurements. A procedure for implementing neural network detection of group test objects is proposed. The obtained results can be used when developing hardware and software for optoelectronic tracking systems.

关键词： hyperspectral image convolutional network detector group object external trajectory measurements

来源：评论

学校读者我要写书评

暂无评论

A Lightweight Multiscale-Multiobject Deep Segmentation Architecture for UAv-Based Consumer Applications

引用

IEEE TRANSACTIONS ON CONSUMER ELECTRONICS 2024年第1期70卷 3740-3753页

作者： Behera, Tanmay Kumar Bakshi, Sambit Khan, Muhammad Attique Albarakati, Hussain Mobarak Natl Inst Technol Rourkela Dept Comp Sci & Engn Rourkela 769008 India HITEC Univ Dept Comp Sci Taxila 47080 Pakistan Lebanese Amer Univ Dept Comp Sci & Math Beirut 11022801 Lebanon Umm Al Qura Univ Coll Comp & Informat Syst Comp & Network Engn Dept Mecca 24382 Saudi Arabia

Smart UAvs have been developed under the consumer Internet of Drone Things (CIoDTs) framework to improve the quality of service (QoS) for several commercial and consumer applications. Artificial intelligence (AI)-inspired algorithms are employed here to make these remote sensing devices more intelligent and agile to perform the task most effectively. However, these AI-based techniques may suffer from obtaining the required feature space, leading to poor performance of the AI system. Thus, to address these intrinsic issues, this manuscript presents a lightweight multiscale convolutional neural network (MCNN), which segments the UAv-captured aerial images based on their object classes. The proposed framework is based on an encoder-decoder architecture developed using depthwise separable convolution. The encoder module here involves extracting the crucial scale-invariant features that are upsampled at the decoder stage to generate the predicted segmented images. Additionally, the proposed architecture is validated using two datasets: the prepared NITRDrone dataset and the Urban Drone Dataset (UDD). Thus, the proposed method achieves an IoU score of 0.877 and 0.802 with the NITRDrone and UDD, respectively. Thus, the proposed lightweight architecture helps enhance the efficiency of these intelligent mobile edge aerial devices for several critical consumer applications demanding real-time data processing.

关键词： Task analysis Remote sensing Computer architecture Autonomous aerial vehicles Convolutional neural networks Consumer electronics Computer vision vHR remote sensing aerial image multiscale CNN semantic segmentation mobile edge devices

来源：评论

学校读者我要写书评

暂无评论

Nonlinear SvM Classifier based Online Attendance System for Facial Recognition and Mask Detection

Nonlinear SVM Classifier based Online Attendance System for ...

引用

2024 International Conference on Knowledge Engineering and Communication systems, ICKECS 2024

作者： Devi, M. varshana Surya, K. A. Jai Meethashree, E. Keerthivasan, v.S. Saraswathy, S. K. S. Rangasamy College of Technology Computer Science and Engineering Tiruchengode637 215 India Sri Sai Ram Engineering College Computer Science and Business Systems Chennai600044 India

ISBN: (纸本)9798350359688

This research presents a novel method for monitoring attendance by creating an online system that combines facial recognition technology with machine learning algorithms to detect masks. The main objective of this project is to provide a streamlined attendance management system that can be accessed through web browsers, removing the requirement for the installation of specialised software. The technology utilises facial biometrics to identify users, storing profiles with facial picture samples in a centralised online database. Before implementing facial recognition, the system goes through a model training phase where Support vector Machines (SvM) are used to create a strong recognition model. Moreover, synthetic data is utilised to train the algorithm in recognising individuals who are wearing facial masks. The server-side application is developed using Python and utilises the OpenCv library for image processing. The web interfaces are managed through PHP, while the database is handled using MySQL. The combination of Python and PHP scripting enables immediate processing on web servers, guaranteeing availability from any device through a web browser. The experimental findings indicate that the face recognition system achieved an accuracy rate of around 81.8%, while the mask detection system achieved an accuracy rate of 80%. These results were obtained using a pre-trained model. In summary, this system provides a whole solution for effective and safe online attendance management in many fields. © 2024 IEEE.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Plant Leaf Diseases Detection Using KNN Classifier 9

Plant Leaf Diseases Detection Using KNN Classifier

引用

9th International Conference on Advanced Computing and Communication systems, ICACCS 2023

作者： Gurunathan, v. Sathiya Priya, T. Dhanasekar, J. Ishwarya Niranjana, M. Suganya, S. Dr. Mahalingam College of Engineering and Technology Department of ECE Pollachi India Sri Eshwar College of Engineering Department of ECE Coimbatore India Hindusthan Institute of Technology Department of ECE Coimbatore India

ISBN: (纸本)9798350397376

In the agricultural area today, identifying Leaf Disease is very challenging. A significant portion of the economy is dependent on agricultural output, thus when a disease is incorrectly identified, there will be a significant loss in crop yield and the economic value of food. The cost of traditional procedures is quite low, but they need human labour to visually inspect the plant's leaf patterns and identify the disease, as well as more time and laborious effort from farmers. Software-based methods for detecting leaf illnesses involve a significant amount of work, expertise in leaf diseases, and longer processing times for the diagnosis. Based on the KNN Classifier, we have suggested image processing for identifying leaf disease. The procedure for identifying a disease includes image preprocessing, contrast enhancement, RGB conversion, feature extraction, segmentation, and classification using K-nearest neighbour algorithms. The quality of the splint samples is improved by using Histogram Equalization, which is initially applied to the samples of leaves after they have been shrunk to 256x256 pixels and image-preprocessed. The instructional characteristics of the splint samples are valued using a matrix called the Grey Level Co-occurrence Matrix. Using machine learning techniques akin to K-Nearest Neighbor, the characteristics are categorized (K- NN). K- NN is used to evaluate the proposed model's quality. © 2023 IEEE.

关键词： Graphic methods

来源：评论

学校读者我要写书评

暂无评论

Performance Evaluation of Multiwavelet Transform for Single image Dehazing 3rd

Performance Evaluation of Multiwavelet Transform for Single ...

引用

3rd EAI International Conference on Cognitive Computing and Cyber Physical systems, IC4S 2022

作者： Sankar, M. Ravi Srinivas, P. Praveena, v. Bhavani, D. Suseela, M. Sri Uma Srinivas, Y. Rao, Ch. venkateswara Department of ECE Sasi Institute of Technology and Engineering Tadepalligudem India Department of ECE SRKR Engineering College Bhimavaram534202 India Department of ECE Vishnu Institute of Technology Bhimavaram534202 India

ISBN: (纸本)9783031289743

images captured in poor lighting conditions (such haze, fog, mist, or smog) have a lower level of visibility because air particles deflect light. Single picture dehazing techniques can restore clarity to a single hazy image. Even though Dark channel prior (DCP) has been the most used method for image dehazing algorithms, it has poor picture quality and requires a lot of adjustments for real time applications such as;computer vision, object detection. However, previous studies such as;Dark Channel Prior (DCP) on image dehazing agonises from a gigantic processing time and sleaze of the original image. In addition, the dimensions of the images have a significant effect on the performance of the dehazing algorithms. Hence, the main motive of this work is to minimize the computational overload either by reducing the size of the image or by accurate transmission map estimation without compromising the image quality. Hence, this work is focused on reducing the computational time for image dehazing through multilevel discrete (Haar) wavelet transform, in which the image dimensions has been reduced without degrading quality. The performance measures of proposed algorithm have been analysed in terms of PSNR, MMSE and SSIM and compared with existing DCP algorithm. The simulation results proven that, the computational time for proposed algorithm has been reduced by 90% when compared to DCP without degrading image quality. © 2023, ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering.

关键词： Wavelet transforms

来源：评论

学校读者我要写书评

暂无评论

A Comparison Between CPU and GPU Computing in DNN-Based DoA Estimation

A Comparison Between CPU and GPU Computing in DNN-Based DoA ...

引用

UNIfied Conference of International Workshop on Defence Applications of Multi-Agent systems, DAMAS 2023, International Conference on Maintenance Engineering, IncoME-v 2023, International conference on the Efficiency and Performance Engineering Network, TEPEN 2023

作者： Kokkinis, Georgios Ahmed, Qasim Z. Sambell, Alistair Chochliouros, Ioannis P. Lazaridis, Pavlos I. Zaharis, Zaharias D. School of Computing and Engineering University of Huddersfield HuddersfieldHD1 3DH United Kingdom Marousi151 24 Greece School of Electrical and Computer Engineering Aristotle University of Thessaloniki Thessaloniki Greece

ISBN: (纸本)9783031494208

Deep Learning algorithms have recently become accessible to modern Telecommunication systems due to the advancement in both hardware and software technology. The massively parallel computational tasks of Artificial Intelligence (AI) training are now directly performed in Graphic processing Units (GPUs), which contain many processing cores. This paper tests the improvement in computing time for Deep Neural Network (DNN) architectures that solve the Direction of Arrival (DoA) estimation problem as a classification task by utilizing the Nvidia CUDA framework. The training and testing data are generated by receiving signal information from a simulated antenna array. The input of the DNN is a 64 × 64 image with two channels, whereas the problem is modelled as a multi-label classification task. The purpose of this research is to observe the acceleration that the GPU provides in the DoA estimation classification problem. The project includes benchmarks from a wide range of computing systems and the results demonstrate the potential of DNN-based DoA estimators as a modern technology that can be deployed on available commercial hardware. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

A Deep Learning-based Radiomics Approach for COvID-19 Detection from CXR images using Ensemble Learning Model 36

A Deep Learning-based Radiomics Approach for COVID-19 Detect...

引用

36th IEEE International Symposium on Computer-Based Medical systems (CBMS)

作者： Costa, Marcus v. L. de Aguiar, Erikson J. Rodrigues, Lucas S. Ramos, Jonathan S. Traina, Caetano, Jr. Traina, Agma J. M. Univ Sao Paulo Inst Math & Comp Sci ICMC BR-13566590 Sao Carlos SP Brazil Fed Univ Rondonia UNIR BR-76801059 Porto Velho RO Brazil

ISBN: (纸本)9798350312249

Medical image analysis plays a major role in aiding physicians in decision-making. Specifically in detecting COvID-19, Deep Learning (DL) and radiomic approaches have achieved promising results separately. However, DL results are hard to interpret/visualize, and the radiomic approach encompasses successive steps, such as image acquisition, image processing, segmentation, feature extraction, and analysis. In this paper, we integrate DL with radiomic approaches, aiding in detecting COvID-19. We use DL models to extract 128 relevant deep radiomic features to assess COvID-19 from several image sources of 392 representative chest X-ray (CXR) exams. We avoid successive radiomic steps by employing DL (transfer learning) from imagenet's vGG-16, ResNet50v2, and DenseNet201 networks. We considered a set of Machine Learning (ML) algorithms to further validate our results, providing an ensemble model to detect COvID-19. Our experimental results show that our approach achieved 95% AUC using 128 relevant features from DenseNet201. Conversely, our ensemble model presented 91% AUC, indicating that deep learning-based radiomics could increase binary classification performance in a real scenario. In addition, we highlight that our approach can be adapted to create other DL-based radiomics tools. For reproducibility, we made our code available at https://***/usmarcv/CBMS-DL-based-radiomics.

关键词： Deep Learning Deep Radiomics Deep Features Radiomics Medical images

来源：评论

学校读者我要写书评

暂无评论

Deep Learning-Based image Defect Detection and Removal in Manufacturing

Deep Learning-Based Image Defect Detection and Removal in Ma...

引用

Digital Optical Technologies 2023

作者： Gapon, N. voronin, v. Zhdanova, M. Sizyakin, R. Semenishchev, E. Ilyukhin, Y. Center for Cognitive Technology and Machine Vision Moscow State University of Technology «STANKIN» Moscow Russia Don State Technical University Rostov-on-Don Russia

ISBN: (数字)9781510664586

ISBN: (纸本)9781510664579

image inpaintnig in textile manufacturing is a new emerging research topic in preprocessing for jacquard CAD systems. One of the most important aspects of a jacquard CAD system is the simulation of the appearance of a jacquard texture during inference. Jacquard image inpainting has become an indispensable process for the Jacquard CAD system. Jacquard image reconstruction is designed to restore a damaged image with missing information, so it is necessary to determine which parts of the image need to be repaired. Thus, this task includes two processing stages: the detection of defects and their recovery. This article presents a two-stage approach that combines new and traditional algorithms for detecting defects and repairing damaged areas. The first stage is a defect detection method based on a convolutional autoencoder (U-Net). The second stage is image inpainting based on exemplar-based concepts and the anisotropic gradient. Our system quantitatively outperforms state-of-the-art methods regarding reconstruction accuracy in the benchmark. © 2023 SPIE.

关键词： image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Optimizing Convolutional image processing on RISC-v Using Branch Prediction on FPGA

Optimizing Convolutional Image Processing on RISC-V Using Br...

引用

IEEE Region 10 International Conference TENCON

作者： Mekala Bindu Bhargavi Sujay Krishnan Subramanian Sri Parameswaran Soumya J Department of EEE BITS Pilani-Hyderabad Campus Hyderabad India Department of Electrical and Computer Engineering University of Sydney Australia

ISBN: (数字)9798350350821

ISBN: (纸本)9798350350838

This paper explores the enhancement of convolutional image processing on a RISC-v based architecture through the implementation of branch prediction (BP) techniques. Con-volution, a critical operation in image processing, is essential for tasks such as edge detection, blurring, and feature extraction. Efficient execution of convolution operations, particularly in a RISC-v architecture, requires effective BP to minimize delays caused by conditional operations. This study presents the development of a 5-stage pipelined 32-bit RISC-v processor, integrated with a 2-bit saturating counter branch predictor using a Branch History Table (BHT). The efficiency of the RISC-v processor was evaluated by synthesizing and implementing it using Xilinx vivado on the Kintex- 7 KC705 Evaluation Platform and testing with a convolution algorithm written in assembly language. Experimental results demonstrate that optimizing BP significantly improves the execution efficiency of convolution operations, making it a promising approach for advanced image processing applications.

关键词： Machine learning algorithms Convolution image processing image edge detection Prediction algorithms Throughput Real-time systems History Assembly Testing

来源：评论

学校读者我要写书评

暂无评论

Latency-aware Spatial-wise Dynamic Networks 36

Latency-aware Spatial-wise Dynamic Networks

引用

36th Conference on Neural Information processing systems (NeurIPS)

作者： Han, Yizeng Yuan, Zhihang Pu, Yifan Xue, Chenhao Song, Shiji Sun, Guangyu Huang, Gao Tsinghua Univ Dept Automat BNRist Beijing Peoples R China Peking Univ Sch Elect Engn & Comp Sci Beijing Peoples R China

ISBN: (纸本)9781713871088

Spatial-wise dynamic convolution has become a promising approach to improving the inference efficiency of deep networks. By allocating more computation to the most informative pixels, such an adaptive inference paradigm reduces the spatial redundancy in image features and saves a considerable amount of unnecessary computation. However, the theoretical efficiency achieved by previous methods can hardly translate into a realistic speedup, especially on the multi-core processors (e.g. GPUs). The key challenge is that the existing literature has only focused on designing algorithms with minimal computation, ignoring the fact that the practical latency can also be influenced by scheduling strategies and hardware properties. To bridge the gap between theoretical computation and practical efficiency, we propose a latency-aware spatial-wise dynamic network (LASNet), which performs coarse-grained spatially adaptive inference under the guidance of a novel latency prediction model. The latency prediction model can efficiently estimate the inference latency of dynamic networks by simultaneously considering algorithms, scheduling strategies, and hardware properties. We use the latency predictor to guide both the algorithm design and the scheduling optimization on various hardware platforms. Experiments on image classification, object detection and instance segmentation demonstrate that the proposed framework significantly improves the practical inference efficiency of deep networks. For example, the average latency of a ResNet-101 on the imageNet validation set could be reduced by 36% and 46% on a server GPU (Nvidia Tesla-v100) and an edge device (Nvidia Jetson TX2 GPU) respectively without sacrificing the accuracy. Code is available at https://***/LeapLabTHU/LASNet.

关键词： Graphics processing unit

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：