检索结果-内蒙古大学图书馆

Paradigm shift from artificial neural networks (ANNs) to deep Convolutional neural networks (DCNNs) in the field of medical image processing

引用

EXPERT SYSTEMS WITH applications 2024年 244卷

作者： Abut, Serdar Okut, Hayrettin Kallail, K. James Siirt Univ Dept Comp Engn TR-56100 Siirt Turkiye Univ Kansas Dept Off Res Sch Med 1010 N Kansas Wichita KS 67214 USA Univ Kansas Dept Populat Hlth Sch Med 1010 N Kansas Wichita KS 67214 USA Univ Kansas Dept Internal Med Sch Med 1010 N Kansas Wichita KS 67214 USA

images and other types of unstructural data in the medical domain are rapidly becoming data-intensive. Actionable insights from these complex data present new opportunities but also pose new challenges for classification or segmentation of unstructural data sources. Over the years, medical problems have been solved by combining traditional statistical methods with image processing methods. Both the increase in the size of the data and the increase in the resolution are among the factors that shape the ongoing improvements in artificial intelligence (AI), particularly concerning deep learning (DL) techniques for evaluation of these medical data to identify, classify, and quantify patterns for clinical needs. At this point, it is important to understand how artificial neural networks (ANNs), which are an important milestone in interpreting big data, transform into Deep Convolutional neural networks (DCNNs) and to predict where the change will go. We aimed to explain the needs of these stages in medical image processing through the studies in the literature. At the same time, information is provided about the studies that lead to paradigm shift and try to solve the image related medical problems by using DCNNs. With the increase in the knowledge of medical doctors on this subject, it will be possible to look at the solution of new problems in computer science from different perspectives.

关键词： Deep Convolutional neural networks Medical image processing artificial neural networks Feature Extraction

来源：评论

学校读者我要写书评

暂无评论

SqueezeCapsNet: enhancing capsule networks with squeezenet for holistic medical and complex images

引用

MULTIMEDIA TOOLS AND applications 2023年第1期83卷 2823-2852页

作者： Adu, Kwabena Walker, Joojo Mensah, Patrick Kwabena Ayidzoe, Mighty Abra Opoku, Michael Boateng, Samuel Univ Energy & Nat Resources Dept Comp Sci & Informat Sunyani 00233 Ghana Univ Elect Sci & Technol China Sch Informat & Software Engn Chengdu 610054 Sichuan Peoples R China

Early diagnosis of patients' disease is crucial since it helps doctors and patients devise a treatment plan. Therefore, recognizing medical images using artificial intelligence-based deep learning techniques has recently increased. Capsule Network (CapsNet) has promising methods in visual tasks due to its ability to keep a high relationship of spatial information compared to convolutional neural networks (CNNs). However, CapsNet faces a critical problem with a complex image background that limits its performance. The traditional CapsNet adopts a standalone convolution (SC) as a feature extractor, Softmax function for normalization of coupling coefficient, and dynamic routing procedure to allow active capsules to perform predictions leading to activation of high-level capsules. The SC is not an effective feature extractor, and SoftMax impedes capsules from distributing optimal coupling coefficient during routing. This paper proposes a CapsNet architecture called SqueezeCapsNet that integrates SqueezeNet and CapsNet to achieve effective feature extraction and fewer parameters. A new squash function named parametric squash function (PSF) was proposed to reduce non-informative capsules and promote discriminative capsules. To the best of our knowledge in literature, we are the first to integrate SqueezeNet into CapsNet. We evaluate our framework on two medical image datasets;Brain tumor and Lung & Colon cancer datasets. Additionally, datasets with varied backgrounds;MNIST, fashion-MNIST, CIFAR-10 were used to evaluate the robustness and generalizability of the model. The SqueezeCapsNet produces 94.85%, 99.76%, 99.87%, 93.49%, and 82.45% on Brain tumor, Lung & Colon Cancer, MNIST, fashion-MNIST, and CIFAR-10 datasets, respectively. Experimental results show that the proposed architecture's compression techniques significantly provide fewer parameters while enhancing stability and accuracy across all the evaluation metrics. Our results show that our method improves CapsNet

关键词： artificial intelligence Capsule networks image processing Medical images Squeeze network

来源：评论

学校读者我要写书评

暂无评论

A collaborative inference strategy for medical image diagnosis in mobile edge computing environment

引用

PEERJ COMPUTER SCIENCE 2025年 11卷 e2708-e2708页

作者： Zhang, Shiqian Cui, Yong Xu, Dandan Lin, Yusong Zhengzhou Univ Sch Comp & Artificial Intelligence Zhengzhou Henan Peoples R China Zhengzhou Univ Collaborat Innovat Ctr Internet Healthcare Zhengzhou Henan Peoples R China Zhengzhou Univ Light Ind Sch Comp & Commun Engn Zhengzhou Henan Peoples R China Zhengzhou Univ Sch Cyber Sci & Engn Zhengzhou Henan Peoples R China

The popularity and convenience of mobile medical image analysis and diagnosis in mobile edge computing (MEC) environments have greatly improved the efficiency and quality of healthcare services, necessitating the use of deep neural networks (DNNs) for image analysis. However, DNNs face performance and energy constraints when operating on the mobile side, and are limited by communication costs and privacy issues when operating on the edge side, and previous edge-end collaborative approaches have shown unstable performance and low search efficiency when exploring classification strategies. To address these issues, we propose a DNN edge-optimized collaborative inference strategy (MOCI) for medical image diagnosis, which optimizes data transfer and computation allocation by combining compression techniques and multi-agent reinforcement learning (MARL) methods. The MOCI strategy first uses coding and quantization-based compression methods to reduce the redundancy of image data during transmission at the edge, and then dynamically segments the DNN model through MARL and executes it collaboratively between the edge and the mobile device. To improve policy stability and adaptability, MOCI introduces the optimal transmission distance (Wasserstein) to optimize the policy update process, and uses the long short-term memory (LSTM) network to improve the model's adaptability to dynamic task complexity. The experimental results show that the MOCI strategy can effectively solve the collaborative inference task of medical image diagnosis and significantly reduce the latency and energy consumption with less than a 2% loss in classification accuracy, with a maximum reduction of 38.5% in processing latency and 71% in energy consumption compared to other inference strategies. In real-world MEC scenarios, MOCI has a wide range of potential applications that can effectively promote the development and application of intelligent healthcare.

关键词： Mobile edge computing Medical imaging Collaborative inference Feature compression Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Intelligent hyperspectral image processing for landmine detection and classification 4

Intelligent hyperspectral image processing for landmine dete...

引用

4th International Workshop of IT-Professionals on artificial Intelligence, ProfIT AI 2024

作者： Sineglazov, Viktor Lesohorskyi, Kyrylo Lytvynenko, Volodymyr National Aviation University Liubomyra Huzara Ave 1 Kyiv03058 Ukraine National Technical University of Ukraine Ihor Sikorsky Kyiv Polytechnic Institute Beresteiskyi Ave 37 Kyiv03056 Ukraine Kherson National Technical University Beryslavs'ke Hwy 24 Kherson Oblast Kherson73008 Ukraine

This work is devoted to a hybrid three-stage approach to hyperspectral image classification to solve the problem of remote landmine detection. A comprehensive overview of landmine pollution and its effects is given. Challenges of landmine detection and classification are highlighted. An overview of existing projects and practical applications is given. The method that utilizes a two-step pre-processing followed by a robust convolution neural network-based feature extractor and classifier is proposed. The proposed method is considered and tested in both batch and real-time scenarios. The results show state-of-the-art accuracy and viability of real-time landmine detection, however, the method is prone to high false positive incidence. © 2024 Copyright for this paper by its authors.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Research on multi-source microstructure image recognition of foam ceramics using convolutional network combine with frequency domain

引用

SCIENTIFIC REPORTS 2025年第1期15卷 1-17页

作者： Yin, Yi Pan, Jianwei Wang, Fang Li, Peihang Cai, Zhen Xu, Xin Wuhan Univ Sci & Technol Sch Comp Sci & Technol Wuhan 430081 Peoples R China Wuhan Univ Sci & Technol State Key Lab Refractories & Met Wuhan 430081 Peoples R China Huazhong Univ Sci & Technol Wuhan Natl Lab Optoelect Wuhan 430074 Peoples R China Univ Durham Comp Sci Durham DH1 3LE England Minist Educ Joint Int Res Lab Refractories & Met Wuhan 430081 Peoples R China

Foam ceramics are widely used in industrial applications due to their unique properties, including high porosity, lightweight, and high-temperature resistance. However, their complex microstructure presents significant challenges for image analysis. Traditional machine learning methods often fall short in capturing both global feature dependencies and detailed representations. To address this, a novel artificial intelligence recognition model, FD-Conv, is proposed, which combines the global information processing capabilities of Transformers with the local feature extraction strengths of convolutional neural networks. Additionally, a frequency domain block detail enhancement mechanism is introduced to improve recognition accuracy. Experimental results demonstrate that the FD-Conv model enhances recognition accuracy by at least 7.6% compared to state-of-the-art methods. Furthermore, the model effectively identifies foam ceramics with varying compositions and formulations and quantifies their microstructural phase characteristics. This research aims to advance the application of foam ceramic microstructure image analysis by improving recognition accuracy, particularly in multi-source microscopic image feature learning and pattern recognition.

关键词： Foam ceramics Multi-source Microstructure images FD-Conv Frequency domain block

来源：评论

学校读者我要写书评

暂无评论

Single image super-resolution with self-organization neural networks and image laplace gradient operator

引用

MULTIMEDIA TOOLS AND applications 2022年第8期81卷 10607-10630页

作者： Ahmadian, Khodabakhsh Reza-Alikhani, Hamid-reza Islamic Azad Univ Dept Elect & Comp Engn Mahshahr Branch Mahshahr Iran Natl Univ Dept Elect & Comp Engn Tafresh Branch Tafresh Iran

At present, artificial neural networks have received wide applications in the field of image processing and image resolution because of their fast algorithm implementation and their high accuracy. Learning-based super-resolution methods used stochastic computation in their algorithms, leading to a manual and experimental adjustment of the regularization parameter to the solving imaging system model problem. In this paper, we present a new hybrid algorithm for low-resolution image enhancement, whose parameters are automatically adjusted by the training data and, in contrast to other super-resolution methods, do not require regular adjustment parameters. The method is a hybrid method that includes self-organizing maps as a preprocessor, the k-nearest neighbor algorithm as a classifier, and the Laplace operational edge detection operator as an edge extractor. We built a single external dictionary using a combination of low-resolution and high-resolution feature patches and then train our proposed network. Subsequently, we reconstruct the high-resolution image by Converting the low-resolution input image to feature patch vectors. Then for each vector, find the matching neuron in the network and retrieve all the vectors that belong to it. Then we train the k-nearest neighbor algorithm with these vectors plus the input vector and find the best vector most similar to the input vector and reconstruct our high super-resolution image. The proposed image super-resolution method presents in practical experiments better results with better resolution and quality than many traditional and state-of-the-art methods, both visually compared with each other using human and computational benchmarks to compare the quality of the image super-resolution algorithms. The proposed image enhancement method is best for reconstructing high-resolution images that need high-frequency details and sharp edges with a smooth slope of image objects in their structures.

关键词： image super-resolution Self-organizing map K-nearest neighbor image enhancement Laplace gradient operator image processing

来源：评论

学校读者我要写书评

暂无评论

Corrosion Modelling Using Convolutional neural networks: A Brief Overview

引用

Journal of Bio- and Tribo-Corrosion 2022年第3期8卷 1-8页

作者： Idusuyi, Nosa Samuel, Oluwatosin Joshua Olugasa, Temilola Taiwo Ajide, Olusegun Olufemi Abu, Rahaman Department of Mechanical Engineering University of Ibadan P.O. Box 22133 Oyo State Nigeria

Convolutional neural Network (CNN) is a type of artificial neural network which is trained using image data. This network architecture consists of a convolutional base and a dense head which helps in classification tasks. This trained networks is then used to solve complex problems like determining the magnitude of damage caused by corrosion. Typically the design of a CNN model would involve image collection, pre-processing, feature extraction and analysis. This paper presents a brief overview of various applications of CNN-based models to corrosion in selected industries. The use of transfer learning to build corrosion CNN models is also discussed. When they are combined with recursive algorithms, the application of CNN models to pinpoint exact locations where corrosion occurs is discussed. From the works reviewed, CNN models can be applied when limited data are available using the freeze transfer learning approach. Convolutional neural networks have shown promising applications for corrosion classification purposes with accuracies above 80%. © 2022, The Author(s), under exclusive licence to Springer Nature Switzerland AG.

关键词： Corrosion

来源：评论

学校读者我要写书评

暂无评论

LoockMe: An Ever Evolving artificial Intelligence Platform for Location Scouting in Greece 1

引用

24th International Conference on Engineering applications of neural networks (EANN)

作者： Trivizakis, Eleftherios Aidonis, Vassilios Pezoulas, Vassilios C. Goletsis, Yorgos Oikonomou, Nikolaos Stefanis, Ioannis Chondromatidou, Leoni Fotiadis, Dimitrios I. Tsiknakis, Manolis Marias, Kostas Fdn Res & Technol Hellas FORTH Computat BioMed Lab CBML Iraklion 70013 Greece Fdn Res & Technol Hellas FORTH Biomed Res Inst BRI Ioannina 45110 Greece Univ Ioannina Dept Mat Sci & Engn Unit Med Technol & Intelligent Informat Syst Ioannina 45110 Greece Univ Ioannina Dept Econ Lab Business Econ & Decis Ioannina 45110 Greece Hellen Mediterranean Univ Dept Elect & Comp Engn Iraklion 71410 Greece

ISBN: (数字)9783031342042

ISBN: (纸本)9783031342035;9783031342042

LoockMe is an artificial intelligence-powered location scouting platform that combines deep learning image analysis, cutting-edge machine learning natural language processing (NLP) for automated content annotation, and intelligent search. The platform's objective is to label input images of local landscapes, and/or any other assets that regional film offices want to expose to those interested in identifying potential locations for the film production industry. The deep learning-based image analysis achieved high classification performance with an AUCscore of 99.4%. Moreover, the state-of-the-art machine learning NLP module enhances the platform's capabilities by analyzing text descriptions of the locations and thus allowing for automated annotation, while the intelligent search engine combines image analysis with NLP to extract relevant context from available data. The proposed artificial intelligence platform has the potential to substantially assist asset publishers and revolutionize the location scouting process for the film production industry in Greece.

关键词： artificial intelligence deep learning transfer learning natural language processing location scouting film production search engine

来源：评论

学校读者我要写书评

暂无评论

Dress-up: deep neural framework for image-based human appearance transfer

引用

MULTIMEDIA TOOLS AND applications 2023年第15期82卷 23151-23178页

作者： Ghodhbani, Hajer Neji, Mohamed Qahtani, Abdulrahman M. Almutiry, Omar Dhahri, Habib Alimi, Adel M. Univ Sfax Natl Engn Sch Sfax ENIS REs Grp Intelligent Machines REGIM Lab BP 1173 Sfax 3038 Tunisia Natl Sch Elect & Telecommun Sfax Technopk BP 1163 Sfax 3018 Tunisia Taif Univ Coll Comp & Informat Technol Dept Comp Sci POB 11099 Taif 21944 Saudi Arabia King Saud Univ Coll Appl Comp Sci Riyadh Saudi Arabia Univ Johannesburg Fac Engn & Built Environm Dept Elect & Elect Engn Sci Johannesburg South Africa

The fashion industry is at the brink of radical transformation. The emergence of artificial Intelligence (AI) in fashion applications creates many opportunities for this industry and make fashion a better space for everyone. Interesting to this matter, we proposed a virtual try-on interface to stimulate consumers purchase intentions and facilitate their online buying decision process. Thus, we present, in this paper, our flexible person generation system for virtual try-on that aiming to treat the task of human appearance transfer across images while preserving texture details and structural coherence of the generated outfit. This challenging task has drawn increasing attention and made huge development of intelligent fashion applications. However, it requires different challenges, especially in the case of a wide divergences between the source and target images. To solve this problem, we proposed a flexible person generation framework called Dress-up to treat the 2D virtual try-on task. Dress-up is an end-to-end generation pipeline with three modules based on the task of image-to-image translation aiming to sequentially interchange garments between images, and produce dressing effects not achievable by existing works. The core idea of our solution is to explicitly encode the body pose and the target clothes by a pre-processing module based on the semantic segmentation process. Then, a conditional adversarial network is implemented to generate target segmentation feeding respectively, to the alignment and translation networks to generate the final output results. The novelty of this work lies in realizing the appearance transfer across images with high quality by reconstructing garments on a person in different orders and looks from simlpy semantic maps and 2D images without using 3D modeling. Our system can produce dressing effects and provide significant results over the state-of-the-art methods on the widely used DeepFashion dataset. Extensive evaluations show th

关键词： artificial intelligence Outfit generation Garment interchange Virtual try-on Semantic segmentation

来源：评论

学校读者我要写书评

暂无评论

FPGA-based reflection image removal using cognitive neural networks

引用

Applied Nanoscience (Switzerland) 2023年第3期13卷 2539-2553页

作者： Saptalakar, Bairu K. Latte, Mrityunjaya V Department of Electronics and Communication Engineering SDM College of Engineering and Technology Karnataka Dharwad580002 India Department of Electronics and Communication Engineering JSS Academy of Technical Education Karnataka Bengaluru560060 India

There is an enormous increase in the resource usage and certain process is required to satisfy the user requirement. Thus, there is a process integration on IoT and data analytics which paves the way for the smart city development. The energy management plays a crucial role based on the processing of image dataset and optimization using the artificial intelligence. To increase the visibility of the image based on the reflection concept, cognitive models are used by relocating the clear glass images. The above process is used widely in computer vision applications as it is ill nature and makes the additional precursors more challenging. Then, eliminating reflections problem is considered for various heuristic observations or other assumptions and it fulfills in practical conditions. In this paper, we generalize the assumptions for issues of reflection based on usage of different information or impose new limitations elimination. The image elimination helps in effective energy utilization by managing the resource through CNN, i.e., optimization techniques of cognitive networks. To overcome the various optimization functions, long computational time required and reflection removal methods, i.e., conventional have not guaranteed their performance. It helps in developing the smart cities as energy plays a significant role as the propose system is integrated with IoT and data analytics. As FPGA system is deployed, it can screen the images by providing effectively the sustainable policies with energy efficient platform through data compression. In the performance analysis, the proposed algorithm is effective as compared with current approaches using CNN methods. Parallel processing of images in Digital Signal processing tools will create degradation of the images and it is difficult to achieve high performance, so the implementation is done using FPGA. © 2022, King Abdulaziz City for Science and Technology.

关键词： Field programmable gate arrays (FPGA)

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：