检索结果-内蒙古大学图书馆

Three-branch neural network for No-Reference Quality assessment of Pan-Sharpened images

ENGINEERING applications OF artificial INTELLIGENCE 2025年 139卷

作者： Stepien, Igor Oszust, Mariusz Rzeszow Univ Technol Doctoral Sch Al Powstancow Warszawy 12 PL-35959 Rzeszow Poland Rzeszow Univ Technol Dept Comp & Control Engn Wincentego Pola 2 PL-35959 Rzeszow Poland

Pan-Sharpening (PS) techniques aim to enhance the spatial resolution of low-resolution multispectral images by leveraging data from high-resolution panchromatic images. Their comparison typically relies on the quality assessment of the resulting Full-Resolution (FS) pan-sharpened images. However, in the absence of a reference image, a dedicated No-Reference (NR) method must be employed. Therefore, this paper introduces a novel approach called the Three-Branch neural Network for No-Reference Quality Assessment of Pan- Sharpened images (TBN-PSI). The network consists of three subnetworks designed for perceptual processing of image channels, featuring shared extraction of low-level features and high-level semantics. Extensive experimental evaluation demonstrates the superiority of the approach over the state-of-the-art NR PS image quality assessment methods, using six datasets containing diverse satellite images that span urban areas, green vegetation, and water scenarios. Specifically, TBN-PSI outperforms the compared methods by 4% to 9% in terms of Spearman's Rank-Order Correlation Coefficient (SRCC), Pearson's Linear Correlation Coefficient (PLCC), and Kendall's Rank Correlation Coefficient (KRCC) between the obtained scores and those of three representative full-reference methods.

关键词： Remote sensing Pan-sharpening image quality assessment neural networks Three-branch neural network Pan-sharpening quality prediction

来源：评论

学校读者我要写书评

暂无评论

Optical Character Recognition of Persian Epigraphs by Using artificial neural networks in MATLAB

Optical Character Recognition of Persian Epigraphs by Using ...

引用

2024 International conference on artificial Intelligence and Quantum Computation-Based Sensor applications, ICAIQSA 2024

作者： Khan, Sumaiyya Bardekar, Aashish Sipna College of Engineering & Technology Dept. of Computer Science and Engineering Amravati India

ISBN: (纸本)9798331517953

Persian, also known as Farsi, was once a dominant language and cultural influence worldwide. Today, it is nearly extinct in the Indian subcontinent. However, for centuries, Farsi played a major role in high culture and administration across the subcontinent. Persian characters are complex, with various features such as ascenders, descenders, vertical lines, and curves. Automatic translation of Persian characters from inscriptions into English is currently unavailable. The process of deciphering and translating inscribed text in epigraphy can be time-consuming. This paper specifically focuses on identifying Persian epigraphs. First, the inscription image is cleaned up, and images containing Persian characters are obtained. Using preprocessing techniques, the images are then prepared for input to artificial neural networks. The artificial neural networks are trained to effectively recognize various patterns of Persian characters, as they can learn complex patterns and adapt to changes in images, resulting in high accuracy in Persian character recognition. Finally, the proposed method's performance is evaluated using accuracy, precision, and speed metrics. The results demonstrate that the method accurately and rapidly detects Persian characters in various images. This research has the potential to advance Persian character recognition and image processing, enabling broader applications such as converting image documents into text, character recognition in documents, and text translation with Persian image content. © 2024 IEEE.

关键词： MATLAB

来源：评论

学校读者我要写书评

暂无评论

artificial Intelligence Techniques for Medical image Segmentation: A Technical Overview and Introduction to Advanced applications 4th

Artificial Intelligence Techniques for Medical Image Segment...

引用

4th International conference on Digital Technologies and applications (ICDTA)

作者： Sabbar, Hanan Silkan, Hassan Abbad, Khalid Chouaib Doukkali Univ Comp Sci Dept LAROSERI Lab El Jadida Morocco Sidi Mohamed Ben Abdellah Univ Sci Dept Intelligent Syst & Applicat Lab Fes Morocco

ISBN: (纸本)9783031686597;9783031686603

Medical image segmentation plays a crucial role in enhancing diagnostic, treatment, and research applications within the medical field, increasingly relying on sophisticated artificial intelligence (AI) technologies. This survey explores various AI architectures employed in medical image segmentation, including Deep neural networks, Encoder-Decoder networks, Attention-Based networks, and hybrid models that integrate features from multiple architectures. It methodically assesses these models' performances, focusing on their effectiveness across different medical imaging types. The analysis delves into each technique's specific advantages and limitations, offering valuable insights into their practical applications in medical diagnostics and treatment planning. The comparative study highlights the differences in accuracy, processing time, and the ability to manage complex image structures among the methods reviewed. It also explores the synergistic benefits of hybrid models, which are shown to achieve superior segmentation results. By providing a comprehensive comparison of AI-driven segmentation techniques, the paper significantly enhances our understanding of how these technologies can be optimized and implemented effectively in medical image analysis. The ultimate goal is to guide future research and the practical application of AI in medical image segmentation, aiming to develop more efficient and precise diagnostic tools.

关键词： Segmentation Medical image artificial intelligent Deep learning Computer vision

来源：评论

学校读者我要写书评

暂无评论

Spam Email image Detection Using Convolution neural Network and Convolutional Block Attention Module 1

引用

International conference on artificial Intelligence: Theory and applications (AITA)

作者： Jamal, Md Asif Kumar, Pradeep Maulana Azad Natl Urdu Univ Dept Comp Sci & Informat Technol Hyderabad India

ISBN: (数字)9789819984794

ISBN: (纸本)9789819984787;9789819984794

image spam is continually a popular area of research. Cyberspace is under attack from many different directions. Spam that contains text embedded in an image is known as "image spam". The rise in online conversation through email has globally contributed to the increasing rate of spam email relatively. First started text spam then now a new challenge in few years image spam which has been a major problem in the field of computing. Various machine learning techniques are used to classify image spam based on a large number of attributes retrieved from the image. Most existing image spam filtering systems use features created by hand and time-consuming machine learning approaches. Convolution neural networks (CNNs) are commonly utilized in image-processing, classification, and feature extraction applications due to their outstanding results. In this research, we use a two CNN model built using deep learning methods to analyze image spam;one is without attention, and next one is with an attention module that time. We use the convolutional block attention module (CBAM);this module attention only spams area of the image and activities of the better performance. Our proposed method achieved very competitive performance-99.42% accuracy on the image Spam Hunter (ISH) dataset and state-of-the-art performance;in this paper, we used the ISH dataset.

关键词： image spam Deep learning Convolution neural network (CNN) Convolutional block attention module (CBAM)

来源：评论

学校读者我要写书评

暂无评论

EPU: An Energy-Efficient Explainable AI Accelerator With Sparsity-Free Computation and Heat Map Compression/Pruning

引用

IEEE JOURNAL OF SOLID-STATE CIRCUITS 2024年第3期59卷 830-841页

作者： Kim, Junsoo Han, Seunghee Ko, Geonwoo Kim, Ji-Hoon Lee, Changha Kim, Taewoo Youn, Chan-Hyun Kim, Joo-Young Korea Adv Inst Sci & Technol Sch Elect Engn Daejeon 34141 South Korea

Deep neural networks (DNNs) have recently gained significant prominence in various real-world applications such as image recognition, natural language processing, and autonomous vehicles. However, due to their black-box nature in system, the underlying mechanisms of DNNs behind the inference results remain opaque to users. In order to address this challenge, researchers have focused on developing explainable artificial intelligence (AI) algorithms. Explainable AI aims to provide a clear and human-understandable explanation of the model's decision, thereby building more reliable systems. However, the explanation task differs from well-known inference and training processes as it involves interactions with the user. Consequently, existing inference and training accelerators face inefficiencies when processing explainable AI on edge devices. This article introduces explainable processing unit (EPU), the first hardware accelerator designed for explainable AI workloads. The EPU utilizes a novel data compression format for the output heat maps and intermediate gradients to enhance the overall system performance by reducing both memory footprint and external memory access. Its sparsity-free computing core efficiently handles the input sparsity with negligible control overhead, resulting in a throughput boost of up to 9.48x. It also proposes a dynamic workload scheduling with a customized on-chip network for distinct inference and explanation tasks to maximize internal data reuse hence reducing external memory access by 63.7%. Furthermore, the EPU incorporates point-wise gradient pruning (PGP) that can significantly reduce the size of heat maps by a factor of 7.01x combined with the proposed compression format. Finally, the EPU chip fabricated in a 28 nm CMOS process achieves a remarkable heat map generation rate of 367 frames/s for ResNet-34 while maintaining the state-of-the-art area and energy efficiency of 112.3 GOPS/mm(2) and 26.55 TOPS/W, respectively.

关键词： artificial intelligence Heat maps Training Convolutional neural networks Task analysis Labeling Semantics Convolutional neural network (CNN) deep neural network (DNN) explainable artificial intelligence (XAI) multiple DNN acceleration neural processing unit (NPU)

来源：评论

学校读者我要写书评

暂无评论

Unlocking the Potential of Spiking neural networks: Understanding the What, Why, and Where

引用

IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL SYSTEMS 2024年第5期16卷 1648-1663页

作者： Wickramasinghe, Buddhi Chowdhury, Sayeed Shafayet Kosta, Adarsh Kumar Ponghiran, Wachirawit Roy, Kaushik Purdue Univ Sch Elect & Comp Engn W Lafayette IN 47907 USA

Spiking neural networks (SNNs) are a promising avenue for machine learning with superior energy efficiency compared to traditional artificial neural networks (ANNs). Recent advances in training and input encoding have put SNNs on par with state-of-the-art ANNs in image classification. However, such tasks do not utilize the internal dynamics of SNNs fully. Notably, a spiking neuron's membrane potential acts as an internal memory, merging incoming inputs sequentially. This recurrent dynamic enables the networks to learn temporal correlations, making SNNs suitable for sequential learning. Such problems can also be tackled using ANNs. However, to capture the temporal dependencies, either the inputs have to be lumped over time (e.g., Transformers);or explicit recurrence needs to be introduced [e.g., recurrent neural networks (RNNs) and long-short-term memory (LSTM) networks], which incurs considerable complexity. To that end, we explore the capabilities of SNNs in providing lightweight solutions to four sequential tasks involving text, speech, and vision. Our results demonstrate that SNNs, by leveraging their intrinsic memory, can be an efficient alternative to RNNs and LSTMs for sequence processing, especially for certain edge applications. Furthermore, SNNs can be combined with ANNs (hybrid networks) synergistically to obtain the best of both worlds in terms of accuracy and efficiency.

关键词： Neurons Training Task analysis Encoding Membrane potentials Computational modeling Recurrent neural networks Automatic speech recognition neuromorphic computing optical flow estimation spiking neural networks (SNNs)

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). 8

Fire Incendie Detection A compartive study of Convolutional ...

引用

8th IEEE International conference on image and Signal processing and their applications, ISPA 2024

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

ISBN: (纸本)9798350309249

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. © 2024 IEEE.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Detection of corrosion on steel structures using an artificial neural network

引用

STRUCTURE AND INFRASTRUCTURE ENGINEERING 2023年第12期19卷 1860-1871页

作者： Khayatazad, Mojtaba Honhon, Matthias De Waele, Wim Univ Ghent Fac Engn & Architecture Dept Electromech Syst & Met Engn SOETE Lab B-9052 Zwijnaarde Belgium SIM VZW Zwijnaarde Belgium

image based-corrosion detection has become a widespread practice for steel structures, but fine-tuning their model parameters is time-consuming. Alternatively, convolutional neural networks (CNNs) can also be trained fast and automatically, but they demand a huge training dataset. In this paper, a corrosion detection approach based on an artificial neural network (ANN) whose training dataset size is less than 0.1% of that of typical CNNs is introduced. The input layer of the proposed ANN consists of textural and color properties. In the present work, different color spaces and textural properties are examined for their impact on the robustness of the ANN. Results reveal that the best color channels can be achieved by combining CIE L*u*v* and YUV color spaces. Moreover, energy is selected as the best texture feature with respect to the ANN robustness. The proposed ANN outperforms an available image processing algorithm from the perspective of both speed and accuracy. In conclusion, this ANN can be used for actual applications after a fast and straightforward training step.

关键词： Steel structures image-based corrosion detection artificial neural networks deep hidden layers color space texture metric

来源：评论

学校读者我要写书评

暂无评论

Emerging memristors and applications in reservoir computing

引用

Frontiers of physics 2024年第1期19卷 219-239页

作者： Hao Chen Xin-Gui Tang Zhihao Shen Wen-Tao Guo Qi-Jun Sun Zhenhua Tang Yan-Ping Jiang School of Physics and Optoelectronic Engineering Guangdong University of TechnologyGuangzhou Higher Education Mega CenterGuangzhou 510006China

Recently, with the emergence of ChatGPT, the field of artificial intelligence has garnered widespread attention from various sectors of society. Reservoir Computing (RC) is a neuromorphic computing algorithm used to analyze time-series data. Unlike traditional artificial neural networks that require the weight values of all nodes in the trained network, RC only needs to train the readout layer. This makes the training process faster and more efficient, and it has been used in various applications, including speech recognition, image classification, and control systems. Its flexibility and efficiency make it a popular choice for processing large amounts of complex data. A recent research trend is to develop physical RC, which utilizes the nonlinear dynamic and short-term memory properties of physical systems (photonic modules, spintronic devices, memristors, etc.) to construct a fixed random neural network structure for processing input data to reduce computing time and energy. In this paper, we introduced the recent development of memristors and demonstrated the remarkable data processing capability of RC systems based on memristors. Not only do they possess excellent data processing ability comparable to digital RC systems, but they also have lower energy consumption and greater robustness. Finally, we discussed the development prospects and challenges faced by memristors-based RC systems.

关键词： reservoir computing memristor resistive switching artificial synapse neuromorphic computing

来源：评论

学校读者我要写书评

暂无评论

Interpolation Based Distributed Learning 32

Interpolation Based Distributed Learning

引用

32nd IEEE Signal processing and Communications applications conference (SIU)

作者： Akdis, Mehmet Amasyali, Mehmet Fatih Yildiz Tekn Univ Bilgisayar Muhendisligi Bolumu Istanbul Turkiye

ISBN: (纸本)9798350388978;9798350388961

artificial neural networks are trained with gradient descent-based algorithms in a complex parameter error space. As a result of the training, a point (local minimum) with a small error on the training set is obtained in this space. However, due to the random processes of the training algorithms, different local minima are reached even if starting from very close points using the same data set. When linear interpolation is made between these reached points, it is seen that the error values of the obtained points are high. In this study, it has been shown that this situation does not occur among models trained in parallel with different data after a common training, and that very successful results can be achieved when training continues from the best point obtained by interpolation. In other words, instead of training sequentially with all the data, more successful results could be obtained by adding a parallel process to the training process in approximately the same time. This approach, which does not require information sharing between models throughout the parallel process, differs from the literature in this respect and offers an innovative training method for artificial neural networks.

关键词： linear interpolation distributed training artificial neural networks gradient descent parameter-loss space

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：