检索结果-内蒙古大学图书馆

Solving Optimization Problems Through Fully Convolutional Networks: An Application to the Traveling Salesman Problem

引用

IEEE TRANSACTIONS ON systems MAN CYBERNETICS-systems 2021年第12期51卷 7475-7485页

作者： Ling, Zhengxuan Tao, Xinyu Zhang, Yu Chen, Xi Zhejiang Univ Coll Control Sci & Engn Hangzhou 310027 Peoples R China Zhejiang Univ Coll Control Sci & Engn Key Lab Ind Control Technol Hangzhou 310027 Peoples R China

In the new wave of artificial intelligence, deep learning is impacting various industries. As a closely related area, optimization algorithms greatly contribute to the development of deep learning. But the reverse applications are still insufficient. Is there any efficient way to solve certain optimization problem through deep learning? The key is to convert the optimization to a representation suitable for deep learning. In this article, a traveling salesman problem (TSP) is studied. Considering that deep learning is good at image processing, an image representation method is proposed to transfer a TSP to an image. Based on samples of a ten city TSP, a fully convolutional network (FCN) is used to learn the mapping from a feasible region to an optimal solution. The training process is analyzed and interpreted through stages. A visualization method is presented to show how an FCN can understand the training task of a TSP. Once the training is completed, no significant effort is required to solve a new TSP and the prediction is obtained on the scale of milliseconds. The results show good performance in finding the global optimal solution. Moreover, the developed FCN model has been demonstrated on TSP's with different city numbers, proving excellent generalization performance.

关键词： Urban areas Optimization Deep learning Neural networks Feature extraction image resolution Heuristic algorithms Deep learning fully convolutional network (FCN) mapping optimization traveling salesman

来源：评论

学校读者我要写书评

暂无评论

Machine learning and image processing to monitor strain and tensile forces with mechanochromic sensors

引用

EXPERT systems WITH APPLICATIONS 2023年 212卷

作者： de Castro, Lucas D. C. Scabini, Leonardo Ribas, Lucas C. Bruno, Odemir M. Oliveira Jr, Osvaldo N. Univ Sao Paulo Sao Carlos Inst Phys BR-13566590 Sao Carlos SP Brazil Sao Paulo State Univ Inst Biosci Humanities & Exact Sci BR-15054000 Sao Jose Do Rio Preto SP Brazil

A computer vision (CV) system is proposed for real-time prediction of strain by monitoring the color-changing feature of mechanochromic sensors. Pictures of the sensors subjected to calibration tensile tests were treated with standard image processing methods and analyzed using supervised machine learning (ML) algorithms. Visual strain sensing was demonstrated by linear regression models capable of learning a relation between the applied strain and the reflected structural color. The ElasticNet regression model provided the highest accuracy in the strain prediction task, with a remarkable performance in monitoring real-time strain variation of sensors during a tensile-relaxion cycle. Using calibration curves, the predicted strain can also be employed for estimating the tensile force applied on the mechanochromic sensors. Taken together these results point to potential intelligent systems for noninvasive in-situ visual monitoring of deformations and tensions.

关键词： Sensors Mechanochromic Computer vision Machine learning image processing

来源：评论

学校读者我要写书评

暂无评论

Design of computing granularity configurable processor based on RISC-V extended instruction 16

Design of computing granularity configurable processor based...

引用

16th IEEE International Conference on Solid-State and Integrated Circuit Technology, ICSICT 2022

作者： He, Bin Yu, Ning-Mei Xu, Meng Wang, Xing-Jia Xi'an University of Technology School of Automation and Information Engineering Xi'an710000 China

ISBN: (数字)9781665469067

ISBN: (纸本)9781665469067

Aiming at the problems of long operation time and poor flexibility of embedded processor, this paper proposes a structure of configurable execution unit based on RISC-V extended instruction set for computing granularity. The structure sets up basic parallel computing units and configuration registers, which can realize a large number of parallel operations such as multiplication, addition, or higher abstractions similar to convolutional neural networks by software configuration. This paper completed the design of this structure and integrated it into a four-stage pipelined RISC-V processor. The white blood cell images collected by lensless system are used to test various image algorithms. The experimental results show that the processor speed designed in this paper is 5.95 times to 22.21 times higher than that of the ordinary pipeline processor, and the number of instructions required is reduced by 1.16 times to 7.49 times. Therefore, the processor has a certain improvement in processing speed and flexibility, which is of great significance for the implementation of various algorithms in embedded systems. © 2022 IEEE.

关键词： Pipeline processing systems

来源：评论

学校读者我要写书评

暂无评论

TinyML Applied in Hyperspectral image Classification on COTS Microcontroller

TinyML Applied in Hyperspectral Image Classification on COTS...

引用

Brazilian Symposium on Computing System Engineering (SBESC)

作者： João Victor Santos Hütner Felipe Viel Cesar A. Zeferino Eduardo Augusto Bezerra Laboratory of Embedded and Distributed Systems - LEDS University of Vale do Itajaí - UNIVALI Itajaí SC Brazil SpaceLab Federal University of Santa Catarina - UFSC Florianópolis SC Brazil

ISBN: (数字)9798331509002

ISBN: (纸本)9798331532857

Commercial off-the-shelf (COTS) microcontrollers used in remote sensing are increasingly being used because of the great versatility of application and development provided by this type of component. There has been a significant increase in missions using CubeSats, mainly due to the greater accessibility and flexibility of the COTS equipment used, as well as lower development costs related to this type of equipment. However, CubeSats have main constraints in power, size, and data transmission capacity. As a solution, numerous works use onboard data processing to compress or classify the captured data to reduce the use of storage memory and the amount of data sent to Earth stations. One of the types of data collected by remote sensing satellites are hyperspectral images, which despite being computationally challenging for classification in microcontrollers. Several algorithms are capable of doing hyperspectral image classification even that constrained by low-power microcontrollers with the advances in deep learning. In this sense, the aim and contribution of this study was to demonstrate and explore the feasibility and performance of implementing a deep learning model on a COTS microcontroller for hyperspectral image classification. We developed a convolutional neural network (CNN) architecture adapted for microcontrollers using the TinyML concept with the TensorFlow Lite library to demonstrate the feasibility and performance of deep learning in COTS microcontroller. The results for the overall accuracy of the embedded model were from 82.57 % to 95.81 % for three different literature datasets. The dissipated power was 105 mW for all dataset samples, with the execution time per sample varying from 12.14 ms to 93.18 ms, and the total energy consumed varying from 4.09 mJ to 7.52 mJ for the entire dataset tested.

关键词： Deep learning Adaptation models Satellites Microcontrollers Tiny machine learning Computational modeling Convolutional neural networks CubeSat Hyperspectral imaging image classification

来源：评论

学校读者我要写书评

暂无评论

High-Utilization, High-Flexibility Depth-First CNN Coprocessor for image Pixel processing on FPGA

引用

IEEE TRANSACTIONS ON VERY LARGE SCALE INTEGRATION (VLSI) systems 2021年第3期29卷 461-471页

作者： Colleman, Steven Verhelst, Marian Katholieke Univ Leuven Dept Elect Engn ESAT MICAS B-3001 Leuven Belgium

Recently, CNNs are increasingly exploited for pixel processing tasks, such as denoising, which opens up new challenges due to the increased activation and operation count. This article presents a CNN coprocessor architecture to solve these challenges on field-programmable gate array (FPGA) through four main contributions. First, the I/O communication between the host processor and the FPGA is reduced to a minimum using a depth-first (DF) principle. Three new DF approaches are presented. Second, to ensure high throughput, the increased parallelization opportunities of the proposed line-based DF operation are analyzed. Third, introducing programmability to the compute array is introduced to enable a broad deployment while maintaining high utilization of the available multipliers digital signal processings (DSPs), independently of the kernel dimensions and without control of the host processor. This is in contrast with many state-of-the-art FPGA implementations, focusing on only one algorithm and/or one kernel topology. Fourth, a model is built to investigate the influence of architecture parameters and show the benefits of DF. The scalable design can be deployed on a wide range of FPGAs, maintaining 78%-93% DSP utilization across all algorithms (denoising, optical flow, depth estimation, segmentation, and super-resolution) and FPGA platforms. Up to 695 GOPS is achieved on a Zynq XCZU9EG board, matching state-of-the-art performance with a more flexible design. The throughput is compared with other pixel processing architectures on FPGA.

关键词： Field programmable gate arrays Computer architecture Signal processing algorithms Task analysis Kernel Coprocessors Throughput Convolutional neural network (CNN) depth-first (DF) field-programmable gate array (FPGA) flexible processor design high throughput pixel processing

来源：评论

学校读者我要写书评

暂无评论

Ultrathin arrayed camera for high-contrast near-infrared imaging

引用

OPTICS EXPRESS 2021年第2期29卷 1333-1339页

作者： Kim, Kisoo Jang, Kyung-Won Bae, Sang-In Kim, Hyun-Kyung Cha, Younggil Ryu, Jae-Kwan Jo, Yong-Jin Jeong, Ki-Hun Dept Bio & Brain Engn 291 Daehak Ro Daejeon 34141 South Korea KAIST Inst Hlth Sci & Technol 291 Daehak Ro Daejeon 34141 South Korea Korea Adv Inst Sci & Technol KAIST 291 Daehak Ro Daejeon 34141 South Korea LIG Nexl Co Ltd Unmanned Robot Syst Lab Seongnam 13488 South Korea

We report an ultrathin arrayed camera (UAC) for high-contrast near infrared (NIR) imaging by using microlens arrays with a multilayered light absorber. The UAC consists of a multilayered composite light absorber, inverted microlenses, gap-alumina spacers and a planar CMOS image sensor. The multilayered light absorber was fabricated through lift-off and repeated photolithography processes. The experimental results demonstrate that the image contrast is increased by 4.48 times and the MTF 50 is increased by 2.03 times by eliminating optical noise between microlenses through the light absorber. The NIR imaging of UAC successfully allows distinguishing the security strip of authentic bill and the blood vessel of finger. The ultrathin camera offers a new route for diverse applications in biometric, surveillance, and biomedical imaging. (C) 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

关键词： image processing algorithms image quality Imaging systems Medical imaging Microlens arrays Scanning electron microscopy

来源：评论

学校读者我要写书评

暂无评论

Monitoring PHB production in Synechocystis sp. with hyperspectral images

引用

WATER SCIENCE AND TECHNOLOGY 2022年第1期86卷 211-226页

作者： Rodriguez Lorenzo, Francisco Placer Lorenzo, Miguel Herrero Castilla, Luz Alvarez Rodriguez, Juan Antonio Iglesias, Sandra Gomez, Santiago Fernandez Montenegro, Juan Manuel Rueda, Estel Diez-Montero, Ruben Garcia, Joan Gonzalez-Flo, Eva AIMEN Ctr Aplicac Laser Robot & Control Unit Poligono Ind Cataboi SUR PPI 2Sect 2Parcela 3 O Porrino 36418 Pontevedra Spain AIMEN Environm Technol Unit Ctr Aplicac Laser Poligono Ind Cataboi SUR PPI 2Sect 2Parcela 3 O Porrino 36418 Pontevedra Spain Univ Politecn Catalunya BarcelonaTech Escola Engn Barcelona Est EEBE Dept Civil & Environm Engn GEMMA Grp Environm Engn & Microbiol Av Eduard Maristany 16Bldg C5-1 E-08019 Barcelona Spain Univ Politecn Catalunya UPC Dept Civil & Environm Engn GEMMA Grp Environm Engn & Microbiol C Jordi Girona 1-3Bldg D1 E-08034 Barcelona Spain Univ Cantabria Dept Water & Environm Sci & Technol GIA Grp Environm Engn Santander Spain

Microalgae wastewater treatment systems have the potential for producing added-value products. More specifically, cyanobacteria are able to accumulate polyhydroxybutyrates (PHBs), which can be extracted and used for bioplastics production. Nonetheless, PHB production requires proper culture conditions and continue monitoring, challenging the state-of-the-art technologies. The aim of this study was to investigate the application of hyperspectral technologies to monitor cyanobacteria population growth and PHB production. We have established a ground-breaking measurement method able to discern spectral reflectance changes from light emitted to cyanobacteria in different phases. All in all, enabling to distinguish between cyanobacteria growth phase and PHB accumulation phase. Furthermore, first tests of classification algorithms used for machine learning and image recognition technologies had been applied to automatically recognize the different cyanobacteria species from a complex microbial community containing cyanobacteria and microalgae cultivated in pilot-scale photobioreactors. We have defined three main indicators for monitoring PHB production: (i) cyanobacteria specific-strain density, (ii) differentiate between growth and PHB-accumulation and (iii) chlorosis progression. The results presented in this study represent an interesting alternative for traditional measurements in cyanobacteria PHB production and its application in pilot-scale PBRs. Although not directly determining the amount of PHB production, they would give insights on the undergoing processes.

关键词： chlorosis cyanobacteria hyperespectral image processing PHB photobioreactor

来源：评论

学校读者我要写书评

暂无评论

Real-Time Embedded Object Tracking with Discriminative Correlation Filters Using Convolutional Features 1

引用

18th International Symposium on Applied Reconfigurable Computing

作者： Danilowicz, Michal Kryjak, Tomasz AGH Univ Sci & Technol Dept Automat Control & Robot Comp Vis Lab Embedded Vis Syst Grp Krakow Poland

ISBN: (数字)9783031199837

ISBN: (纸本)9783031199820;9783031199837

Object tracking is an essential element of visual perception systems. It is used in advanced video surveillance systems (AVSS), autonomous vehicles, robotics, and many more. For applications such as autonomous robots, the system must be implemented on some embedded platform with limited computing performance and power. Furthermore, sufficiently fast response is required from the tracking system in order to perform some real-time tasks. Discriminative Correlation Filter (DCF) based tracking algorithms are popular for such applications, as they offer state-of-the-art performance while not being too computationally complex. In this paper, an FPGA implementation of the DCF tracking algorithm using convolutional features is presented. The ZCU104 board is used as a platform, and the performance is evaluated on the VOT2015 dataset. In contrast to other implementations that use HOG (Histogram of Oriented Gradients) features, this implementation achieves better results for 64 x 64 filter size while being able to potentially operate at higher speeds (over 467 fps per scale).

关键词： Discriminative correlation filter Object tracking FPGA Real-time image processing

来源：评论

学校读者我要写书评

暂无评论

Lightweight scale-focused pyramid network for remote sensing image object detection

Lightweight scale-focused pyramid network for remote sensing...

引用

Natural Language processing (ICNLP), International Conference on

作者： Daxiang Li Zhan Ji Ying Liu Yao Tang School of Communication and Information Engineering Xi'an University of Posts and Telecommunications Xi’an China

ISBN: (数字)9798350349115

ISBN: (纸本)9798350349122

Lightweight target detection algorithms are crucial for real-time processing of remote sensing satellite images. However, due to the large size of remote sensing images and high computational costs, there are difficulties in dealing with significant changes in target scales. Existing lightweight models tend to prioritize detection speed at the expense of detection accuracy. To address this issue, we propose a lightweight scale-focused pyramid network(LSFPN). Firstly, we design a lightweight BaseNeck module to reconstruct the network's neck part, eliminating complex feature fusion processes and reducing network complexity. Additionally, we design a multi-scale expansion(MSE) module to optimize multi-scale feature extraction and improve the network's multi-scale target detection performance. Lastly, considering the significant scale variations in remote sensing images, we adopt the K-means++ clustering algorithm to re-determine the initial anchor box sizes, further enhancing the accuracy of target detection. Experimental results on the HRSC2016 dataset demonstrate that our method achieves optimal detection accuracy with reduced computational and parameter costs of 37.70% and 59.42% respectively, while improving accuracy by 0.81% with less model complexity.

关键词： Accuracy Computational modeling Object detection Real-time systems Natural language processing Complexity theory Satellite images Neck Remote sensing image reconstruction

来源：评论

学校读者我要写书评

暂无评论

A Low-Complexity Hyperspectral Anomaly Detection Algorithm and Its FPGA Implementation

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2021年 14卷 907-921页

作者： Lei, Jie Yang, Geng Xie, Weiying Li, Yunsong Jia, Xiuping Xidian Univ State Key Lab Integrated Serv Networks Xian 710071 Peoples R China Sci & Technol Electroopt Control Lab Luoyang 471000 Peoples R China Univ New South Wales Sch Engn & Informat Technol Canberra ACT 2600 Australia

On-board real-time anomaly detection has always been a challenging task in hyperspectral imaging analysis as it requires low computational complexity. Most of the existing anomaly detection algorithms inevitably trade off intensive computational complexity for high detection accuracy. This article presents a fast spectral-spatial anomaly detection algorithm with low complexity in hyperspectral images (HSIs) using morphological reconstruction and a simplified guided filter (Fast-MGD). Since the simple filtering techniques are applied, it is therefore feasible to achieve a field programmable gate array (FPGA)-based hardware implementation. More precisely, an effective deeply pipelined acceleration scheme is developed adopting high-level synthesis to support HSIs that are acquired over different scenes with different sizes and spectral bands. Experimental results show strong advantages of the proposed FPGA-based Fast-MGD in processing speed and resource consumption, while a high detection accuracy is remained. Its applicability in on-board real-time processing is demonstrated and verifie.

关键词： Anomaly detection Field programmable gate arrays image reconstruction Feature extraction Hyperspectral imaging Real-time systems Hardware Anomaly detection guided filtering high-level synthesis (HLS) hyperspectral image (HSI) morphological reconstruction real-time implementation

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：