检索结果-内蒙古大学图书馆

3rd International Conference on Communication, Computing and Electronics Systems, ICCCES 2021

作者： Badiger, Manjunatha Mathew, Jose Alex Sahyadri College of Engineering & Management Adyar India Srinivas Institute of Technology Mangalore India

ISBN: (纸本)9789811688614

Deep learning is a subfield of machine learning and artificial intelligence technique. It employs neural network tasks like image processing, computer vision, voice recognition, machine translation, medical information processing, self-driving vehicles, predictive forecasting, robotics and control, cybersecurity, natural language processing, bioinformatics, and countless others. The performance of a neural network is determined by a variety of factors, and activation functions are an essential element in the design of a neural network. The hidden layer’s activation feature defines the extent to which the network model learns the training data set. The type of prognostications made by the model is regulated by the activation functions present at the output layer. This paper presents a comprehensive review of research studies on different activation functions aimed toward deep learning applications. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： neural networks

来源：评论

学校读者我要写书评

暂无评论

Impact of common fabrication errors on the performance of diffractive neural networks 12

Impact of common fabrication errors on the performance of di...

引用

Conference on Holography, Diffractive Optics, and applications XII Part of SPIE/COS Photonics Asia Conference

作者： Goi, Elena Chen, Mengxiang Schoenhardt, Steffen Gu, Min Univ Shanghai Sci & Technol Inst Photon Chips Shanghai 200093 Peoples R China Univ Shanghai Sci & Technol Ctr Artificial Intelligence Nanophoton Sch Opt Elect & Comp Engn Shanghai 200093 Peoples R China

ISBN: (纸本)9781510657038;9781510657021

Optical machine learning has emerged as an important research area that, by leveraging the advantages inherent to optical signals such as parallelism and high speed, paves the way for a future where optical hardware can process information at the speed of light. For the fabrication, in particular of diffractive neural networks, nanoprinting with direct laser writing methods like two-photon nanolithography (TPN) are of great interest, as they allow for the definition of multilayered networks operating in the visible and infrared wavelength regions with nearly continuous phase delays in a single fabrication step. Such diffractive neural networks may have transformative impact on adaptive optics, data processing and sensing and may be crucial in the development of robust and generalized quantitative phase imaging methods(1) with low computational complexity and memory footprint, to be applied, for example, in biological cell- and tissue imaging problems. However, networks printed with the TPN method are potentially afflicted by a number of fabrication errors, which ultimately limit the performance of the networks. Here, we numerically investigate the performance of optical implementations of neural network for complex field data processing in the form of multi-layer nanoscale diffractive neural networks, trained to perform image classification tasks. We discuss the parameter optimization steps for training and the limitations that typical fabrication errors put on the performance of such optical inference systems and analyse the significance of the quasi-continuous phase delay for classification accuracy.

关键词： Diffractive deep neural networks optical deep learning phase retrieval nanofabrication

来源：评论

学校读者我要写书评

暂无评论

Recognizing Indian Classical Dance Forms Using Transfer Learning 4th

Recognizing Indian Classical Dance Forms Using Transfer Lear...

引用

4th International Conference on Machine Intelligence and Signal processing, MISP 2022

作者： Reshma, M.R. Kannan, B. Jagathyraj, V.P. Sabu, M.K. Department of Computer Applications CUSAT Kerala Kochi India School of Management Studies CUSAT Kerala Kochi India

ISBN: (纸本)9789819900466

The rise of cultural tourism in India and massive digitization over the last decade has necessitated preserving Indian art forms. Recent advances in artificial intelligence (AI) have provided the tools and techniques to make these efforts more efficient. The potential research areas involved are digital image processing, computer vision, and machine intelligence. Since researchers have not much explored the scope of the digital preservation of Indian classical dance, this paper contributes a method to classify different Indian classical dance forms. Computerized dance modeling offers a wide range of uses, including cultural conservation, online learning, entertainment, and even remedial treatment. Because the dataset is small, the paper uses transfer learning models for feature extraction and deep neural networks for classification. The proposed method showed a remarkable accuracy of 97.05 percent. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

CNN-based Plant Leaf Disease Identification and Classification for Sustainable Agriculture 6

CNN-based Plant Leaf Disease Identification and Classificati...

引用

6th International Conference on Mobile Computing and Sustainable Informatics, ICMCSI 2025

作者： Srilatha, A. Praveen, P. School of Computer Science & Artificial Intelligence SR University Warangal506371 India

ISBN: (纸本)9798331522667

The rapid spread of leaf-diseases among the large-scale crops or plants will pose a great risk to the agricultural output and crop yield. Earlier detection and proper analysis of foliar disorders are essential for minimizing yield losses as well as forwarding continuous farming trials. This research presents a novel way to find and analyze leaf diseases using various deep learning algorithms. The proposed approach integrates Convolutional neural networks (CNN) in feature extraction for automatically identifying disease patterns from leaf-based images with high accuracy to exploit manual inspection of plant leaves. To achieve higher accuracy, this model is train and tested on a wide variety from diseased & healthy leaf images. This approach surpasses traditional image processing methods, exhibiting high precision and recall in identifying various disease types across multiple crops. The deep learning-based solution offers a scalable, efficient, and reliable tool for farmers and agricultural experts, potentially revolutionizing plant disease management strategies. In addition, it can expand the dataset and enhance real-time detection ability which probably could be applied into user-friendly mobile applications in future work. © 2025 IEEE.

关键词： Accuracy CNN Deep learning image Classification leaf disease detection

来源：评论

学校读者我要写书评

暂无评论

A Comparison of Convolutional neural Network Models for Driver Fatigue Detection 7

A Comparison of Convolutional Neural Network Models for Driv...

引用

7th International Conference on image and Signal processing and their applications, ISPA 2022

作者： Nedjar, Imane Sekkil, Hicham Mohamed Mebrouki, Mahmoud Bekkaoui, Mokhtaria Higher School of Applied Sciences of Tlemcen Biomedical Engineering Laboratory Tlemcen Algeria Higher School of Applied Sciences Second Cycle Department Tlemcen Algeria Tlemcen Algeria

ISBN: (纸本)9781665480420

Road traffic crashes are among the significant risks facing millions of people around the world every day. Driver fatigue is a salient factor in road accidents. However, overcoming this factor has become possible with the use of artificial intelligence. In fact, with the development of technology, industrial companies in the automotive sector are working on intelligent cars capable of identifying the risks and avoiding them. In this work, we propose a method that identifies driver fatigue. First, we established a comparison between 10 models of Convolutional neural networks (CNNs), to classify the state of both mouth and eyes. After selecting the best of them, we compute the percentage of eye closure (PERCLOS) and yawning frequency of mouth (FOM) to conclude the driver state. The proposed method obtained an accuracy of 87.5%. © 2022 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Physical neural network using skyrmion-based spin torque nano-oscillators

Physical neural network using skyrmion-based spin torque nan...

引用

2024 International Conference on Electrical Technology and Automation Engineering, ETAE 2024

作者： Xiong, Shan Liang, Xue Xing, Xiangjun Zhou, Yan School of Physics & Optoelectronic Engineering Guangdong University of Technology Guangdong Guangzhou510006 China Guangdong Provincial Key Laboratory of Information Photonics Technology Guangdong University of Technology Guangdong Guangzhou510006 China School of Science & Engineering The Chinese University of HongKong Guangdong Shenzhen518172 China

Due to physical limitations on the miniaturization of traditional electronic devices, architectures based on emerging principles have become the focus of current research to meet the needs of rapidly developing information technologies in the post-Moore era. Neuromorphic devices hold huge potential for use in future artificial intelligence (AI) chips beyond conventional architectures. Benefiting from a wealth of nonlinear dynamic characteristics of spin torque nano-oscillators (STNOs), studies of neuromorphic computations and their applications based on STNOs are attracting growing attention. In this article, at first, we construct a magnetic skyrmion-based STNO and analyze its characteristics;on this basis, we propose a physical echo state network (ESN) including eight skyrmion-based STNOs, which is utilized to implement an image recognition task. Micromagnetic simulations of the nonlinear response of skyrmion-based STNOs to current pulses imply that such a physical neural network has remarked performance in handwritten digit recognition. The high precision, low energy consumption, and fast processing speed of STNO-based neuromorphic devices are desirable in multitudinous practical applications, possibly leveraging the use of STNO-based physical neural networks in the field of artificial intelligence. © Published under licence by IOP Publishing Ltd.

关键词： Energy utilization

来源：评论

学校读者我要写书评

暂无评论

An Exploration of the Impact of Training Datasets on Deep Learning-Based Building Extraction 6

An Exploration of the Impact of Training Datasets on Deep Le...

引用

6th International Conference on Internet of Things, Automation and artificial Intelligence, IoTAAI 2024

作者： Yang, Shengcheng Song, Chengshun Bao, Liang Zhao, Guangde China Construction Third Engineering Bureau Group Co. LTD Technical Department Xining China

ISBN: (纸本)9798350386974

Building extraction is important in several applications such as urban planning, disaster assessment and navigation systems, and helps to improve the accuracy and application efficiency of spatial data. In recent years, deep learning techniques have progressed significantly in the field of building extraction. However, the quality and diversity of the training dataset are crucial to the model performance. This study aims to investigate the impact of training datasets on deep learning-based building extraction. By comparing datasets of different sizes, quality, and diversity, we evaluate their performance on a deep learning model (UNet) in a building extraction task. The experimental results show that a high-quality and diverse training dataset significantly improves the extraction accuracy and robustness of the model. The deep learning model trained with the richer dataset achieves an overall accuracy of 96.5% for building extraction, which is a 13% improvement in extraction accuracy and a 10% improvement in IoU compared to the model trained on the other dataset. Meanwhile, we found that data enhancement techniques (e.g., image rotation, flipping, and numerical stretching) help improve the generalization ability of the model. The results in this paper not only provide valuable insights into the building extraction task, but also provide methodological guidance for constructing datasets for future deep learning-based remote sensing image processing research. © 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

Improved edge-guided network for single image super-resolution

引用

MULTIMEDIA TOOLS AND applications 2022年第1期81卷 343-365页

作者： Zhao, Jie Chen, Zhenxue Wu, Q. M. Jonathan Li, Xianming Cai, Lei Zhu, Kai Shandong Univ Sch Control Sci & Engn Jinan 250061 Peoples R China Univ Windsor Dept Elect & Comp Engn Windsor ON N9B 3P4 Canada Henan Inst Sci & Technol Sch Artificial Intelligence Xinxiang 453003 Henan Peoples R China

In recent years, deep learning has been successfully applied to image super-resolution. It is still a challenge to reconstruct high-frequency details from low-resolution images. However, many works lack attention to the high-frequency part. We find that edge prior information can be used to extract high-frequency parts and applying soft edges to image reconstruction has achieved great results. Inspired by this, we focus on how to make full use of edge information to generate high-frequency details. We propose an improved edge-guided neural network for single image super-resolution (IEGSR), which makes full use of the edge prior information to reconstruct images with more abundant high-frequency information. For high-frequency information, we propose an edge-net to generate image edges better. For low-frequency information, we propose a global and local feature extraction module (GLM) to reconstruct the texture details. For the fusion of high-frequency information and low-frequency information, we propose a progressive fusion method, which can greatly reduce the number of parameters. Extensive experimental results demonstrate that our method can obtain images with sharper details. Applying our model to the Manga109 test set, the PSNR value of 4 times image super-resolution is as high as 39.02.

关键词： Single image super-resolution Edge prior Feature fusion Convolutional neural network Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Graph convolutional networks of reconstructed graph structure with constrained Laplacian rank

引用

MULTIMEDIA TOOLS AND applications 2022年第24期81卷 34183-34194页

作者： Zhan, Mengmeng Gan, Jiangzhang Lu, Guangquan Wan, Yingying Guangxi Normal Univ Coll Comp Sci & Informat Technol Guilin 541004 Guangxi Peoples R China Massey Univ Sch Nat & Computat Sci Auckland Campus Auckland 0745 New Zealand

Convolutional neural networks (CNNs) have achieved unprecedented competitiveness in text and two-dimensional image data processing because of its good accuracy performance and high detection speed. Graph convolutional networks (GCNs), as an extension of classical CNNs in graph data processing, have attracted wide attention. At present, GCNs often use domain knowledge (such as citation recommendation system, biological cell networks) or artificial created fixed graph to achieve various semi-supervised classication tasks. Poor quality graph will lead to suboptimal results of semi-supervised classification tasks. We propose a more general GCN of reconstructed graph structure with constrained Laplacian rank. First, we use hypergraph to establish multivariate relationships between data. On the basis of the hypergraph, In virtue of Laplacian rank constraint to the graph matrix, we learn a new graph structure which hascconnected components (wherecis the number of classification), and then we construct an ideal graph matrix which is more suitable for the task of semi-supervised classification on GCNs. Finally, the data and the new graph are input GCNs model to get the results of classification. Experiments on 10 different datasets demonstrate that this method is more competitive than the comparison method.

关键词： Graph convolutional networks Adaptive graph Hypergraph Semi-supervised classification Graph structure

来源：评论

学校读者我要写书评

暂无评论

Fire Incendie Detection A compartive study of Convolutional neural networks (CNN) and Support Vector Machines (SVM).

Fire Incendie Detection A compartive study of Convolutional ...

引用

International Conference on image and Signal processing and their applications (ISPA)

作者： Yasmine Guerbai Asma Saibi Department of Electrical Systems Engineering M‘Hammed Bougara University Boumerdès Boumerdès Algeria

ISBN: (数字)9798350309249

ISBN: (纸本)9798350309256

In recent years, the application of artificial intelligence (AI) techniques for fire detection has gained significant attention due to its potential for enhancing early fire detection systems. This study aims to compare the performance of deep learning convolutional neural networks (CNN) and support vector machine (SVM) machine learning algorithms in the context of fire detection. We present a comprehensive analysis and evaluation of the two approaches, highlighting their strengths and weaknesses, and discussing their potential for real-world fire detection applications.

关键词： Support vector machines Deep learning Machine learning algorithms neural networks Signal processing Vectors 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：