检索结果-内蒙古大学图书馆

IEEE International SOC conference

作者： Bongjin Kim School of Electrical and Electronic Engineering Nanyang Technological University Singapore 50 Nanyang Avenue Singapore

ISBN: (数字)9781728134833

ISBN: (纸本)9781728134840

Von Neumann architecture is recently facing a critical challenge with the high demands of energy-efficient computing hardware for a variety of machine learning tasks such as image classification. In particular, battery-operated mobile devices with limited power budget cannot process artificial neural networks (ANNs) with relatively low complexity by using traditional digital circuits and architectures. The key challenge with the traditional Von Neumann architecture is its energy-inefficient data access between memory and processor. Recently, in-memory computing architecture has gained significant attention as an alternative, especially for running mobile artificial intelligence applications. The memory access energy has been drastically reduced by using local in-memory processing elements which directly use the data stored in local memory. To further improve the efficiency, the processor based on mixed-signal circuits instead of conventional digital circuits have recently been actively researched. However, mixed-signal circuits have several critical drawbacks, including nonlinearity, PVT variation, and the overhead of ADC/DAC for interfacing with the external digital domain. In this work, we first review the recent mixed-signal circuits and architectures for in-memory computing using different embedded memories. In addition, we introduce the concept of in-sensor computation for integrating partial computing units in an image-sensor array using low-power mixed-signal circuit techniques.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Supervised texture classification using several features extraction techniques based on ANN and SVM

Supervised texture classification using several features ext...

引用

ACS/IEEE International conference on Computer Systems and applications

作者： Mohammed W. Ashour Mahmoud F. Hussin Khaled M. Mahar College of Engineering & Technology Arab Academy for Science Technology and Maritime Transport Alexandria Egypt

Texture classification is one of the most important clues of visual processing applications .In this paper, we present a comparison between the most two popular supervised texture classification methods based on the feed forward artificial neural Network (ANN) and the multi-class Support Vector Machine (SVM). Five of the most common used features extraction approaches were chosen in order to extract input vectors of different sizes for both classifiers. These approaches are namely gray level histogram, edge detection, and co-occurrence matrices, besides Gabor and Biorthogonal wavelet transformations. Experiments are conducted on two different datasets the first one is engineering surface textures produced by different machining processes, and the second was taken from Brodatz (1966) textures album. The classification accuracy rate is calculated for ANN and SVM in order to measure the efficiency of each technique based on the several features extraction methods. The results show that SVM with its linear and polynomial kernels is higher in classification accuracy and faster in training time.

关键词： Feature extraction Support vector machines Support vector machine classification artificial neural networks Feeds Histograms image edge detection Data engineering Surface texture Machining

来源：评论

学校读者我要写书评

暂无评论

Efficient Hardware Implementations of Deep neural networks: A Survey

Efficient Hardware Implementations of Deep Neural Networks: ...

引用

International conference on Inventive Systems and Control (ICISC)

作者： Sunny Bodiwala Nirali Nanavati Computer Engineering Department Gujarat Technological University Ahmedabad India Computer Engineering Department Sarvajanik College of Engineering and Technology Surat India

ISBN: (数字)9781728128139

ISBN: (纸本)9781728128146

Deep learning develops a significant breakthrough in numerous artificial Intelligence (AI) applications. It is widely used in many state-of-the-art applications such as image classification, speech recognition, object detection and market prediction. However, deep learning models require extensive calculations and high computation power to perform a task. Therefore, there is a need to explore efficient hardware for Deep neural networks (DNNs) without affecting accuracy and its output. Hardware cost is also playing an essential role to affect the DNNs usage in portable or standalone devices. In this paper, we aim to providing detailed survey about recent trends in efficient hardware processing of DNNs. The overview of deep learning approaches, various hardware architectures and key advances in efficient computation of DNNs via reducing hardware size and changing hardware driven algorithms is also analyzed. Furthermore, we have explained the key comparisons, tradeoffs between different hardware platforms and various deep learning techniques for the efficient utilization of resources.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Locating Wood Pith in a Wood Stem Cross Sectional image Using YOLO Object Detection

Locating Wood Pith in a Wood Stem Cross Sectional Image Usin...

引用

International conference on Technologies and applications of artificial Intelligence (TAAI)

作者： Wattanapong Kurdthongmee Korrakot Suwannarat Division of Computer Engineering Walailak University Nakornsithammarat Thailand

Sawmills rely on locating wood pith at the end surface of a wood stem either to automate the process and maximize the yield or to perform wood quality assessment of wood planks. Previously, wood pith location was approximated by applying image processing algorithms along with some geometric relations between the annual rings within a cross-sectional image of wood and its pith. This produces failed detection results if the wood stems deviate from a circular shape. In this paper, we propose employing YOLO (you only look once) object detection to a wood pith identification problem. The Tiny-YOLO network with 23 layers was trained using a transfer learning approach with 345 wood cross-sectional images with pith annotations. Different experiments with variable ratios between the training and testing datasets and training iterations were carried out to produce checkpoints and test against the testing datasets. We found that the minimum loss function which is commonly used to indicate the best trained model in other deep neural networks (DNNs) object detection is not effective for this application. Therefore, we used detection accuracy, to reflect the percentage of detectable piths while inferencing the trained model with its testing dataset and an average distance error, i.e., the average distances between the detected piths and their ground truths, as the accuracy measurement. The experimental result shows that a ratio of training and testing datasets of 70:30, at training iterations of 170,000, gives rise to the maximum detection value of 76.3% with a 16.6-pixel average distance error. Additionally, it produces only 25% average distance error; 10.34/41.76 with 50% standard deviation; 25.33/53.95, compared to the state of the art nonDNN-based pith detection approach.

关键词： Training Testing Object detection Graphics processing units Shape neural networks Standards

来源：评论

学校读者我要写书评

暂无评论

Fusing shrinking and expanding active contour models for robust iris segementation

Fusing shrinking and expanding active contour models for rob...

引用

International Symposium on Signal processing and Its applications (ISSPA)

作者： Kien Nguyen Clinton Fookes Sridha Sridharan Image and Video Research Laboratory Queensland University of Technology Brisbane QLD Australia

The performance of iris recognition systems is significantly affected by the segmentation accuracy, especially in non-ideal iris images. This paper proposes an improved method to localise non-circular iris images quickly and accurately. Shrinking and expanding active contour methods are consolidated when localising inner and outer iris boundaries. First, the pupil region is roughly estimated based on histogram thresholding and morphological operations. Thereafter, a shrinking active contour model is used to precisely locate the inner iris boundary. Finally, the estimated inner iris boundary is used as an initial contour for an expanding active contour scheme to find the outer iris boundary. The proposed scheme is robust in finding exact the iris boundaries of non-circular and off-angle irises. In addition, occlusions of the iris images from eyelids and eyelashes are automatically excluded from the detected iris region. Experimental results on CASIA v3.0 iris databases indicate the accuracy of proposed technique.

关键词： artificial neural networks Legged locomotion Biological system modeling image segmentation Morphology

来源：评论

学校读者我要写书评

暂无评论

Approaches of artificial intelligence in biomedical image processing: A leading tool between computer vision & biological vision

Approaches of artificial intelligence in biomedical image pr...

引用

(Spring) International conference on Advances in Computing, Communication, & Automation (ICACCA)

作者： Shruti Shukla Ashish Lakhmani Ambuj Kumar Agarwal Department of Biomedical Engineering-B. Tech graduated (BME) Babu Banarsi Das-Northern India Engineering College Lucknow Uttar Pradesh Department of Computing Science Engineering-B. Tech Graduate (CSE) Babu Banarsi Das-National Institute Of Technology & Management Lucknow Uttar Pradesh College of Computing Science and Information Technology Teerthankar Mahaveer University Moradabad India

artificial Intelligence or AI is a subfield of computer science, which can be defined as the intelligence exhibited by a machine or a software having a remarkable impact on the field of biology and medicine. Imaging, on the other hand has become an essential component of many fields in medicine, biomedical applications, biotechnology and laboratory research by which images are processed and analysed. Putting together AI and imaging, the tools and techniques of artificial intelligence are useful for solving many biomedical problems and using a computer based equipped hardware software application for understanding images, researchers and clinicians can enhance their ability to study, diagnose, monitor, understand and treat medical disorders. Therefore the main idea behind this research paper is to focus on understanding the artificial intelligence, its concepts and various models available for the segmentation(or classification) of medical images, its applications, advantages and disadvantages and results and more.

关键词： image segmentation artificial intelligence Medical diagnostic imaging artificial neural networks Computational modeling Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Investigating The Role Of Low-Dimensional neural networks In Near-Field Acoustics

Investigating The Role Of Low-Dimensional Neural Networks In...

引用

International conference on Communication Systems and Network Technologies (CSNT)

作者： Nishi Jain V Divya Vani Vijilius Helena Raj Amit Dutt Mohamed I. Habelalmateen Dinesh Kumar Yadav Department of Electrical & Electronics Engineering IES College Of Technology Bhopal M. P. India Institute of Aeronautical Engineering Dundigal Hyderabad Department of Applied Sciences New Horizon College of Engineering Bangalore Lovely Professional University Phagwara The Islamic university Najaf Iraq Knowledge Park II Lloyd Institute of Engineering & Technology Greater Noida Uttar Pradesh

ISBN: (数字)9798350305463

ISBN: (纸本)9798350305470

Noise detection and factory product quality testing are all possible using near-field acoustics. This study examines low-dimensional neural networks for near-field noises. This study examines how artificial neural networks represent and comprehend near-source noises. To determine the advantages and downsides of using these networks. Many near-field acoustics data are utilized to develop and train low-dimensional neural network models for the investigation. networks may detect subtle trends and features in sound data that other approaches overlook. This research evaluates models based on their predicted performance, ease of understanding, and realtime performance. The study shows how low-dimensional neural networks might advance near-field acoustics. Results demonstrate greater noise reduction, quicker signal processing, and more accurate sound source recognition. The study examines how difficult low-dimensional neural networks are to employ in real-world audio applications and what resources are needed. Finally, this study examines low-dimensional neural networks in near-field acoustics, adding to the field's understanding. Acoustics-based systems may improve and be utilized in additional circumstances using the findings. This might enhance environmental, industrial, and other tracking.

关键词： neural networks Noise Noise reduction Signal processing algorithms Acoustics Production facilities Product design

来源：评论

学校读者我要写书评

暂无评论

Classification system for GPR parameters

Classification system for GPR parameters

引用

Proceedings of the International conference on Ground Penetrating Radar, GPR

作者： R. Yelf W. Al-Nuaimy Georadar Research Pty Ltd Coffs Harbour NSW Australia Department of Electrical Engineering and Electronics University of Liverpool Liverpool UK

This paper addresses a problem commonly encountered in the processing and interpretation of GPR data for civil engineering and geotechnical applications. These projects typically involve large data sets composed of multivariate parameters, which need to be interpreted and coded in such a manner as to provide practical solutions to the site engineer. In this paper we propose a system for approaching this problem that has been field tested on many projects. This system uses two numerical descriptors for classifying complex geotechnical parameters derived from GPR field measurements. The first parameter, called GPR Class, is used to code the relative difficulty of excavation, or danger from instability. The second parameter, the GPR Rock Type Descriptor, is used to code the local soil or rock type and to describe the attributes of the material, such as its weathering condition or degree of cementation. The system is a highly efficient for coding large transects of GPR data. The coding system is also suitable for training artificial neural networks. neural networks are being used with increasing frequency for high dimensional problems of regression or classification. These networks, combined with second-order statistical image descriptors, can then be combined to visualise the data in 3 dimensions in a format readily comprehensible to persons without specialist GPR technical knowledge.

关键词： Ground penetrating radar Data engineering Soil measurements artificial neural networks Optical fiber cables Australia Civil engineering System testing Building materials Frequency

来源：评论

学校读者我要写书评

暂无评论

Sign Language Recognition for Deaf and Dumb People Using Convolution neural Network

Journal of ICT Standardization

引用

Journal of ICT Standardization 2022年第3期10卷 411-426页

作者： Herbaz, Nourdine El Idrissi, Hassan Badri, Abdelmajid Faculty of Sciences and Techniques Mohammedia Hassan II University of Casablanca 28830 Morocco

Gesture recognition is an open phenomenon in computer vision, and one of the topics of current interest. Gesture recognition has many applications in the interpretation of sign language in deaf-mutes, one is in human-computer interaction, and the other is in immersive game technology. For this purpose, we have developed a model of image processing recognition of gestures, based on artificial neural networks, starting from data collection, identification, tracking and classification of gestures, to the display of the obtained results. We propose an approach to contribute to the translation of sign language into voice/text format. In this paper, we present a Moroccan sign language recognition system using a Convolutional neural Network (CNN). This system includes an important data set of more than 20 files. Each file contains 1,000 static images of each signal from several different angles that we collected with the camera. Different Sign Language models were evaluated and compared with the proposed CNN model. The proposed system achieved 99.33% and achieved the best performance with an accuracy rate of 98.7%. © 2022 River Publishers.

关键词： Gesture recognition

来源：评论

学校读者我要写书评

暂无评论

Research on Multi-Label Classification of Chinese Text Based on Word-Label Probability and DistilBERT

Research on Multi-Label Classification of Chinese Text Based...

引用

image processing and Computer applications (ICIPCA), IEEE International conference on

作者： Hong Zhao Weiquan Tian Ding An Shan Mi School of Computer Science and Communication Lanzhou University of Technology Lanzhou China CCCC Big Data Technology Co. Ltd. Lanzhou China

ISBN: (数字)9798350360240

ISBN: (纸本)9798350384161

Multi-label text classification is a fundamental task in the field of natural language processing. Currently, there are issues in the Chinese multi-label text classification tasks, such as insufficient extraction of text label features and a lack of learning about the correlations between labels. To address these challenges, we proposed a Multi-label Text Classification Model based on DistilBERT and Word-Label Probabilities (DBWL). The model consists of a metric space module that integrates word-label probabilities and a deep neural network module based on DistilBERT. Firstly, the trained Labeled Latent Dirichlet Allocation (Labeled-LDA) model is utilized to obtain the probabilities of words associated with labels, which are then used to calculate the text mapping vectors and map them into the same metric space as the labels. Similarity measurement is employed to extract features between text-label pairs and label-label pairs. Then, in the deep neural network module, DistilBERT is employed to jointly embed texts and labels, capturing rich semantic information about text labels, followed by feature extraction using TextCNN. Finally, the features extracted from both modules are weighted and fused for label prediction. Experimental results on two benchmark datasets demonstrate that the proposed model outperforms current mainstream multi-label classification models in terms of key performance metrics.

关键词： Computational modeling Text categorization Semantics artificial neural networks Probability Feature extraction Extraterrestrial measurements Vectors Natural language processing Resource management

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：