检索结果-内蒙古大学图书馆

2021 Innovations in Intelligent Systems and applications Conference, ASYU 2021

作者： Sheshulin, Kirill Chuvakov, Aleksandr Ivaschenko, Anton Samara State Technical University Computer Science Department Samara Russia

ISBN: (纸本)9781665434058

The paper explores a problem of artificial neural networks comparative analysis and adaptation while solving a typical task of image analysis and pattern recognition. This article examines four neural networks using different optimizers to solve a problem of pneumonia detection on X-ray images. The study implements convolutional neural networks, as well as the method of their transfer learning. The architecture of neural networks and the results of their work with various metrics that assess the success of this model are presented. The problem of interpretability of the decision made by a neural network can be solved using a special visualization method presented in the study. Advantages of the proposed research are related with a possibility to use the convolutional neural networks in practical applications. © 2021 IEEE.

关键词： Intelligent systems

来源：评论

学校读者我要写书评

暂无评论

Fuzzy Systems and Data Mining vii - Proceedings of FSDM 2021

Fuzzy Systems and Data Mining VII - Proceedings of FSDM 2021

引用

7th International Conference on Fuzzy Systems and Data Mining, FSDM 2021

ISBN: (纸本)9781643682143

The proceedings contain 52 papers. The topics discussed include: how artificial intelligence may impact your job;a deep neural network model for the prediction of major adverse cardiovascular event occurrences in patients with non-ST-elevation myocardial infarction;do reviews influence real estate marketing: the experience combing with natural language processing;ranking of trapezoidal bipolar fuzzy numbers based on a new improved score function;interpretable dual-feature recommender system using reviews;facial expression recognition and image description generation in Vietnamese;hierarchical digital control system performance;and generalized-multiquadric radial basis function neural networks (RBFNs) with variable shape parameters for function recovery.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A genetic programming-based convolutional deep learning algorithm for identifying COVID-19 cases via X-ray images

引用

artificial INTELLIGENCE IN MEDICINE 2023年第1期142卷 102571-102571页

作者： Najaran, Mohammad Hassan Tayarani Univ Hertfordshire Sch Comp Sci Hatfield England

Evolutionary algorithms have been successfully employed to find the best structure for many learning algorithms including neural networks. Due to their flexibility and promising results, Convolutional neural networks (CNNs) have found their application in many image processing applications. The structure of CNNs greatly affects the performance of these algorithms both in terms of accuracy and computational cost, thus, finding the best architecture for these networks is a crucial task before they are employed. In this paper, we develop a genetic programming approach for the optimization of CNN structure in diagnosing COVID-19 cases via X-ray images. A graph representation for CNN architecture is proposed and evolutionary operators including crossover and mutation are specifically designed for the proposed representation. The proposed architecture of CNNs is defined by two sets of parameters, one is the skeleton which determines the arrangement of the convolutional and pooling operators and their connections and one is the numerical parameters of the operators which determine the properties of these operators like filter size and kernel size. The proposed algorithm in this paper optimizes the skeleton and the numerical parameters of the CNN architectures in a co-evolutionary scheme. The proposed algorithm is used to identify covid-19 cases via X-ray images.

关键词： Genetic programming Deep learning Optimization Evolutionary algorithms COVID-19 Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Bearing Fault image Recognition in Strong Noise Based on Optimized CNN Architecture

Bearing Fault Image Recognition in Strong Noise Based on Opt...

引用

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

作者： Qiang Li Haowen Gai Kaijie Hu Northeastern University Shenyang China Tianjin College University of Science and Technology Beijing Tianjin China Zhonghuan Information College Tianjin University of Technology Tianjin China

ISBN: (数字)9798350360240

ISBN: (纸本)9798350384161

To address the problem of low accuracy and poor stability of bearing diagnostic models under strong background noise, a bearing fault image recognition method is proposed that reduces the randomness of the model by avoiding the need for artificial parameterization, which can introduce random factors. This method is based on hyperparameter optimization of the GoogLeNet convolutional neural network model and decision fusion. First of all, two-dimensional wavelet time-frequency variation of the original signal of the bearing vibration to construct an image dataset, one-dimensional classification is into a two-dimensional image problem; Secondly, three lightweight convolutional neural network architectures are selected for the noise immunity test, to get the best noise-resistant network architecture GoogLeNet; Finally, hyper-parameter optimization is performed for the network, comparing the mesh method, the progressive mesh method, and the group optimization algorithm, respectively, and getting The optimal parameters are then analyzed for decision fusion and visualization of the network. To verify the method proposed in this paper, we used the open Case Western Reserve University bearing dataset. The experimental verification demonstrates that the proposed method achieves a correctness rate of 94.33% in decision fusion under noise, with good accuracy and stability.

关键词： Accuracy image recognition Computer architecture Power system stability White noise Stability analysis Robustness Classification algorithms Convolutional neural networks Optimization

来源：评论

学校读者我要写书评

暂无评论

Apply Yolov4-Tiny on an FPGA-Based Accelerator of Convolutional neural Network for Object Detection 2

Apply Yolov4-Tiny on an FPGA-Based Accelerator of Convolutio...

引用

2022 2nd International Conference on artificial Intelligence and Industrial Technology applications, AIITA 2022

作者： Zhang, Fengxi Li, Yuying Ye, Zhihao School of Telecommunications Engineering Xidian University No.266 Xinglong Section of Xi Feng Rd Shaanxi Province Xi'an710126 China School of Artificial Intelligent Xidian University No.266 Xinglong Section of Xi Feng Rd Shaanxi Province Xi'an710126 China

With the continuous expansion of neural Network technology in the artificial intelligence field, for example, image recognition and retrieval, object detection, pixel processing, automatic speech generation, etc., Convolutional neural networks (CNN) and Deep Learning of neural networks (DNN) have made apparent breakthroughs. To improve the inference speed of images, the combination of FPGA-based acceleration and multiple model quantization methods has become one of the most contemporary alternative methods. This paper designed an FPGA-Based acceleration scheme combining software and hardware and effectively applied it to the Yolov4-Tiny object detection model, realizing the accelerated detection inference process from the original 6-7mins to 383ms. First, it chose the static quantization method of fixed-point numbers, fixed the position of the decimal point, and then added Batch Norm between the convolutional layer and the activation function to form a connection structure. Second, it further improved inference speed on an FPGA with a version of ZYNQ-7020 by increasing the bandwidth cap and reducing bandwidth requirements, employing a massive pipeline design. Finally, in the test of the Coco dataset, the plan has completed a substantial acceleration of the average inference speed of the Yolov4-Tiny object detection model from 7.13mins/Picture to 498.89ms/Picture, which has a high application value in the field of object detection. It dramatically improves the inference speed as well as keeps the average accuracy above 0.95. © Published under licence by IOP Publishing Ltd.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Integrating the dynamic optimization strategies of Adam and AdaGrad

Integrating the dynamic optimization strategies of Adam and ...

引用

International Conference on Intelligent Control and Information processing (ICICIP)

作者： Zhixuan Huang Xiaonan Luo Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin University of Electronic Technology Guilin China

ISBN: (数字)9798331516147

ISBN: (纸本)9798331516154

artificial intelligence (AI) has been a key research area since the 1950s, initially focused on using logic and reasoning to create systems that understand language, control robots, and offer expert advice. With the rise of big data and deep learning, AI has advanced in applications like recommendation systems, image recognition, and machine translation, primarily through optimizing loss functions in deep neural networks to improve accuracy and reduce training *** descent is the core optimization method but faces challenges like slow convergence and local minima. To overcome these, algorithms like Momentum, AdaGrad, RMSProp, Adadelta, Adam, and Nadam have been developed, introducing momentum and adaptive learning rates to accelerate convergence. This paper presents a new optimization algorithm that combines the strengths of Adam and AdaGrad, offering better adaptability to different learning rates.

关键词： Training Heuristic algorithms Optimization methods Machine translation Logic artificial intelligence Robots Recommender systems Convergence Intelligent control

来源：评论

学校读者我要写书评

暂无评论

Tool wear and remaining useful life prediction in micro-milling along complex tool paths using neural networks

引用

JOURNAL OF MANUFACTURING PROCESSES 2021年 71卷 679-698页

作者： Bagri, Sumant Manwar, Ashish Varghese, Alwin Mujumdar, Soham Joshi, Suhas S. Indian Inst Technol Dept Mech Engn Mumbai 400076 Maharashtra India

Monitoring tool wear state and prediction of the remaining useful life in micro-milling help avoid down time due to unalarmed failure of the cutting tool. In this regard, investigations based on experiments, and data-based models, along straight tool paths have been reported in the literature. However, industrial applications often require machining along complex tool paths, where the tool failure possibilities increase multifold, have not been studied elaborately in the literature. Therefore, the main objective of this work is to develop a data-based model, which can predict the tool wear state, and remaining useful life of the tool, considering the effect of tool path complexities due to varying tool path radius in micro-milling. The data for developing the model were generated by performing micro-milling experiments under different processing parameters, such as feed, depth of cut, spindle rotation speed, and tool path radius. The tool images were captured in-situ without removing the tool from the spindle, and image binarization and alignment operations were performed to extract corresponding cutting edge wear features, such as diameter reduction as well as the wear of individual cutting edges. The tool wear classification criteria were defined to categorize the tool condition in three regions: initial wear, steadystate wear, and critical wear. To capture and model the complex mechanism of micro tool wear and failure, artificial neural networks, as well as deep belief networks, were implemented to predict the wear state as well as the remaining-useful-life (RUL) of the tools. It was found that the wear rate increased with increasing tool path radii and the larger radii could lead to catastrophic tool failure. The neural-network models gave 93-99% accuracy for the prediction of wear state classification and RUL.

关键词： Micro end-milling Tool path radius Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

A limited-size ensemble of homogeneous CNN/LSTMs for high-performance word classification

引用

neural COMPUTING & applications 2021年第14期33卷 8615-8634页

作者： Ameryan, Mahya Schomaker, Lambert Univ Groningen Artificial Intelligence & Cognit Engn Fac Sci & Engn Groningen Netherlands

The strength of long short-term memory neural networks (LSTMs) that have been applied is more located in handling sequences of variable length than in handling geometric variability of the image patterns. In this paper, an end-to-end convolutional LSTM neural network is used to handle both geometric variation and sequence variability. The best results for LSTMs are often based on large-scale training of an ensemble of network instances. We show that high performances can be reached on a common benchmark set by using proper data augmentation for just five such networks using a proper coding scheme and a proper voting scheme. The networks have similar architectures (convolutional neural network (CNN): five layers, bidirectional LSTM (BiLSTM): three layers followed by a connectionist temporal classification (CTC) processing step). The approach assumes differently scaled input images and different feature map sizes. Three datasets are used: the standard benchmark RIMES dataset (French);a historical handwritten dataset KdK (Dutch);the standard benchmark George Washington (GW) dataset (English). Final performance obtained for the word-recognition test of RIMES was 96.6%, a clear improvement over other state-of-the-art approaches which did not use a pre-trained network. On the KdK and GW datasets, our approach also shows good results. The proposed approach is deployed in the Monk search engine for historical-handwriting collections.

关键词： Coding scheme Ensemble system End-to-end convolutional long short-term memory Connectionist temporal classification

来源：评论

学校读者我要写书评

暂无评论

artificial Intelligence Based Learning Approach for Leaf Disease Identification and Detection 2nd

Artificial Intelligence Based Learning Approach for Leaf Dis...

引用

2nd International Conference on Data Science, Machine Learning and applications, ICDSMLA 2020

作者： Karuna, G. Sahithi, K. Rupa, B. Amani, R. Swaraja, K. Meenakshi, K. Computer Science and Engineering GRIET Hyderabad India Electronics and Communications Engineering GRIET Hyderabad India

ISBN: (纸本)9789811636899

Plants and Crops get diseased due to many reasons. It might be because of diseases of stems, leaves, roots etc. This Paper mainly congregates on leaves. Leaf Disease identification and Detection has many applications for cultivators and farmers to know whether the plant is diseased or not. So that they can retort in dwarf time and decreasing the loss and then can obtain immense profits. This paper mainly focused at learning the disease of plant through leaves. Here, we scrutinize the leaf through image processing and extract features of particular leaf and then utilizing those features as a dataset and done preprocessing and then administering them in artificial Intelligence based learning algorithms like Convolutional neural networks to find disease. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Learning algorithms

来源：评论

学校读者我要写书评

暂无评论

Collaborative Satellite Computing through Adaptive DNN Task Splitting and Offloading

Collaborative Satellite Computing through Adaptive DNN Task ...

引用

IEEE Symposium on Computers and Communications (ISCC)

作者： Shifeng Peng Xuefeng Hou Zhishu Shen Qiushi Zheng Jiong Jin Atsushi Tagami Jingling Yuan School of Computer Science and Artificial Intelligence Wuhan University of Technology China School of Science Computing and Engineering Technologies Swinburne University of Technology Australia KDDI Research Inc. Japan

ISBN: (数字)9798350354232

ISBN: (纸本)9798350354249

Satellite computing has emerged as a promising technology for next-generation wireless networks. This innovative technology provides data processing capabilities, which facilitates the widespread implementation of artificial intelligence (AI)-based applications, especially for image processing tasks involving deep neural network (DNN). With the limited computing resources of an individual satellite, independently handling DNN tasks generated by diverse user equipments (UEs) becomes a significant challenge. One viable solution is dividing a DNN task into multiple subtasks and subsequently distributing them across multiple satellites for collaborative computing. However, it is challenging to partition DNN appropriately and allocate subtasks into suitable satellites while ensuring load balancing. To this end, we propose a collaborative satellite computing system designed to improve task processing efficiency in satellite networks. Based on this system, a workload-balanced adaptive task splitting scheme is developed to equitably distribute the workload of DNN slices for collaborative inference, consequently enhancing the utilization of satellite computing resources. Additionally, a self-adaptive task offloading scheme based on a genetic algorithm (GA) is introduced to determine optimal offloading decisions within dynamic network environments. The numerical results illustrate that our proposal can outperform comparable methods in terms of task completion rate, delay, and resource utilization.

关键词： Satellites Adaptive systems Wireless networks Collaboration artificial neural networks Delays Proposals Resource management Next generation networking Genetic algorithms

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：