检索结果-内蒙古大学图书馆

SPIE 12th International Symposium on Multispectral image processing and Pattern Recognition, MIPPR 2023

作者： Shi, Jiashuo Zhang, Xinyu National Key Laboratory of Science & Technology on Multi-spectral Information Processing Huazhong University of Science & Technology Wuhan430074 China School of Artificial Intelligence & Automation Huazhong University of Science & Technology Wuhan430074 China

ISBN: (数字)9781510675001

ISBN: (纸本)9781510674998

Conventional imaging detection relies on electrical modulation, but it grapples with limitations in speed and adaptability within complex environments. Intelligent imaging detection, capable of manipulating micro-nano-lightfields generated by imaging objective lenses, is essential. The all-optical deep diffractive neural network (D2NN) offers a potential solution, enabling swift spatial information transformation and recognition. However, existing D2NN versions are vulnerable to optical disturbances due to their reliance on computer-based backpropagation. This paper introduces a robust diffractive neural network (SRNN) mathematical model designed to mitigate the impact of structural errors and lightwave frequency shifting without necessitating hardware modifications. The SRNN model incorporates Gaussian noise at various levels during training to emulate the error distribution of the phase mask, augmenting the network's resilience to hardware errors. Weight-Noise-Injection training assists the network in locating an error-insensitive region, thereby enhancing its robustness. © 2024 SPIE.

关键词： Optical data processing

来源：评论

学校读者我要写书评

暂无评论

Learning Enriched Features for Fast image Restoration and Enhancement

引用

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 2023年第2期45卷 1934-1948页

作者： Zamir, Syed Waqas Arora, Aditya Khan, Salman Hayat, Munawar Khan, Fahad Shahbaz Yang, Ming-Hsuan Shao, Ling Incept Inst Artificial Intelligence Abu Dhabi 54114 U Arab Emirates Mohammed BinZayed Univ Artificial Intelligence Abu Dhabi 54114 U Arab Emirates Monash Univeristy Melbourne Vic 3800 Australia Univ Calif Merced Merced CA 95343 USA Google Mountain View CA 94043 USA Terminus Grp Beijing Peoples R China

Given a degraded input image, image restoration aims to recover the missing high-quality image content. Numerous applications demand effective image restoration, e.g., computational photography, surveillance, autonomous vehicles, and remote sensing. Significant advances in image restoration have been made in recent years, dominated by convolutional neural networks (CNNs). The widely-used CNN-based methods typically operate either on full-resolution or on progressively low-resolution representations. In the former case, spatial details are preserved but the contextual information cannot be precisely encoded. In the latter case, generated outputs are semantically reliable but spatially less accurate. This paper presents a new architecture with a holistic goal of maintaining spatially-precise high-resolution representations through the entire network, and receiving complementary contextual information from the low-resolution representations. The core of our approach is a multi-scale residual block containing the following key elements: (a) parallel multi-resolution convolution streams for extracting multi-scale features, (b) information exchange across the multi-resolution streams, (c) non-local attention mechanism for capturing contextual information, and (d) attention based multi-scale feature aggregation. Our approach learns an enriched set of features that combines contextual information from multiple scales, while simultaneously preserving the high-resolution spatial details. Extensive experiments on six real image benchmark datasets demonstrate that our method, named as MIRNet-v2, achieves state-of-the-art results for a variety of image processing tasks, including defocus deblurring, image denoising, super-resolution, and image enhancement. The source code and pre-trained models are available at https://***/swz30/MIRNetv2.

关键词： Multi-scale feature representation dual-pixel defocus deblurring image denoising super-resolution low-light image enhancement and contrast enhancement

来源：评论

学校读者我要写书评

暂无评论

A Comparative Study between artificial neural networks and Optical neural networks 4

A Comparative Study between Artificial Neural Networks and O...

引用

4th International Conference on Current Research in Engineering and Science applications, ICCRESA 2022

作者： Faris, Zainab Murdas, Ibrahim A. Department of Electrical Engineering University of Babylon Babel Iraq

ISBN: (纸本)9798350334944

artificial intelligence has gained widespread interest from people from all walks of life since the big data era arrived, so artificial neural networks (ANNs) are one of the most important areas of artificial intelligence that have a close relationship and an effective role in diverse artificial intelligence applications, data is loaded between memory and processor during the calculation, which reduces the calculation rate and increases the power consumption. Training artificial neural network parameters take a long time and consume a lot of energy. Therefore, it is necessary to use new neural networks that differ from artificial neural networks, called optical neural networks. Optical neural networks can overcome the above shortcomings. Compared with artificial neural networks, they have obvious advantages such as high speed, high bandwidth, high balanced processing, low latency, and low power consumption. The optical neural network can significantly improve the performance of required tasks by taking advantage of the high-speed characteristics. The study compares artificial neural networks and optical neural networks to train networks by nonlinear activation functions to perform image classification tasks. The neural networks are trained and tested with the help of the KMNIST dataset. The backward propagation occurs using the backpropagation algorithm. © 2022 IEEE.

关键词： Electric power utilization

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 29

Collaborative Satellite Computing through Adaptive DNN Task ...

引用

29th IEEE Symposium on Computers and Communications (IEEE ISCC)

作者： Peng, Shifeng Hou, Xuefeng Shen, Zhishu Zheng, Qiushi Jin, Jiong Tagami, Atsushi Yuan, Jingling Wuhan Univ Technol Sch Comp Sci & Artificial Intelligence Wuhan Peoples R China Swinburne Univ Technol Sch Sci Comp & Engn Technol Hawthorn Vic Australia KDDI Res Inc Fujimino Japan

ISBN: (纸本)9798350354249;9798350354232

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： Satellite computing task splitting DNN partitioning task offloading

来源：评论

学校读者我要写书评

暂无评论

Enhancement of Low-Resolution Remote Sensing images Using ANN 6th

Enhancement of Low-Resolution Remote Sensing Images Using AN...

引用

6th International Conference on Communications and Cyber-Physical Engineering, ICCCE 2023

作者： Malini Devi, G. Divya, D. Mounika, J. Sruthi, Ch. Samaikya, D. Department of CSE G. Narayanamma Institute of Technology and Science for Women Telangana Hyderabad India

ISBN: (纸本)9789819971367

Remote-sensing images particularly crucial in wide geographic visual analysis, including change in climate, 3D modelling, and earth’s global observation. As a result of the bad effects of satellite data technology, the transparency and comprehensibility of remote-sensing images are greatly impacted by weather conditions including fog, snowstorms, hurricanes, clouds, and so forth. Many real-world tasks, such surveillance for night time tracking and marine disasters, are made particularly difficult by minimal photography. Remote sensing photos may benefit from improved interpretation and visualization effects by having their brightness increased using image Enhancement (IE) technology. Picture enhancement, picture super resolution, and other image-processing applications have found significant success with artificial neural networks (ANN). The results of image enhancement algorithms such as CNN and ANN are compared and investigated. The approach outcomes also present aesthetic advantages in the sense of noise removal and retaining the integrity of colors and textures. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： neural networks

来源：评论

学校读者我要写书评

暂无评论

What Constitute an Effective Edge Detection Algorithm? 19th

What Constitute an Effective Edge Detection Algorithm?

引用

19th International Conference on Advanced Intelligent Computing Technology and applications (ICIC)

作者： Premaratne, Prashan Vial, Peter Univ Wollongong Sch Elect Comp & Telecommun Engn North Wollongong NSW 2522 Australia

ISBN: (纸本)9789819947416;9789819947423

Edge detection is essential in every aspect of computer vision from vehicle number plate recognition to object detection in images or video. Preprocessing stage of many image processing tasks usually associated with artificial Intelligence are relying on separating multiple objects from each other before accessing any other information. Over the years, researchers have used Sobel, Prewet or Roberts filter and then relying more on robust Canny method. Today, there have been many improvements on the earlier approaches, and now very much rely on Convolutional neural networks to assist in determining effective edges that would assist immensely in object detection. From that perspective, it is quite evident that effective edge detection is all about eventual object detection. With this notion in mind, it is easy to see what methods work and what methods would not achieve the goals. Deep Learning (DL) approaches have been gaining popularity over the years. Do DL algorithms outperform the conventional edge detection algorithms? If they do, is it time for us to forget about the conventional approaches and resort to the new state-of-the art? Are they transparent, when performing poorly? How reliable are they? Do they perform consistently when unknown data are presented? This article will analyze the existing and emerging edge detection methods with a view to determine their usability and limitations in computer vision applications that would undoubtedly advance the field of image processing and computer vision.

关键词： Deep Learning Convolutional neural networks Edge Detection

来源：评论

学校读者我要写书评

暂无评论

Lightweight Deep neural Network for Joint Learning of Underwater Object Detection and Color Conversion

引用

IEEE TRANSACTIONS ON neural networks AND LEARNING SYSTEMS 2022年第11期33卷 6129-6143页

作者： Yeh, Chia-Hung Lin, Chu-Han Kang, Li-Wei Huang, Chih-Hsiang Lin, Min-Hui Chang, Chuan-Yu Wang, Chua-Chin Natl Taiwan Normal Univ Dept Elect Engn Taipei 106 Taiwan Natl Sun Yat Sen Univ Dept Elect Engn Kaohsiung 80021 Taiwan Natl Yunlin Univ Sci & Technol Dept Comp Sci & Informat Engn Touliu 64002 Yunlin Taiwan

Underwater image processing has been shown to exhibit significant potential for exploring underwater environments. It has been applied to a wide variety of fields, such as underwater terrain scanning and autonomous underwater vehicles (AUVs)-driven applications, such as image-based underwater object detection. However, underwater images often suffer from degeneration due to attenuation, color distortion, and noise from artificial lighting sources as well as the effects of possibly low-end optical imaging devices. Thus, object detection performance would be degraded accordingly. To tackle this problem, in this article, a lightweight deep underwater object detection network is proposed. The key is to present a deep model for jointly learning color conversion and object detection for underwater images. The image color conversion module aims at transforming color images to the corresponding grayscale images to solve the problem of underwater color absorption to enhance the object detection performance with lower computational complexity. The presented experimental results with our implementation on the Raspberry pi platform have justified the effectiveness of the proposed lightweight jointly learning model for underwater object detection compared with the state-of-the-art approaches.

关键词： Object detection image color analysis Training Feature extraction Task analysis Distortion Deep learning Convolutional neural networks deep learning lightweight deep model underwater image processing underwater object detection

来源：评论

学校读者我要写书评

暂无评论

Deep fake image Detection Based on Deep Learning Using a Hybrid CNN-LSTM with Machine Learning Architectures as Classifier 6

Deep fake Image Detection Based on Deep Learning Using a Hyb...

引用

6th International Congress on Human-Computer Interaction, Optimization and Robotic applications, HORA 2024

作者： Al-Dulaimi, Omar Alfarouk Hadi Hasan Kurnaz, Sefer Department of Electrical & Computer Engineering Institute of Graduate Studies Altinbas University Istanbul Turkey

ISBN: (纸本)9798350394634

One of the most important and difficult subjects in social communication is detecting deepfake images and videos. Deepfake techniques have developed widely, making this technology quite available and proficient enough so that there is worry about its bad application. Considering this issue, discovering fake faces is very important for ensuring security and preventing sociopolitical issues on a private and general level. Deep learning provides higher performance than typical image processing approaches when it comes to deepfake detection. This work presents construction of an artificial intelligence system, which is capable of detecting deepfake from more than one dataset. This study proposes neural network models based on deep learning using random forest (RF) and support vector machines (SVM) as classifier for deepfake detection. The use of two classifiers (RF) and (SVM) and their combination with a convolutional neural network is the first study of its kind in the field of deepfake detection in images from three open-source datasets (FaceForensics++, FaceAntiSpoofing, and iFakeFaceDB). This proposed method shows an accuracy of 96%, 87% and 52% in iFakeFaceDB, CelebA-Spoof, FaceForensics++ and respectively. © 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

Physics-Constrained Comprehensive Optical neural networks 38

Physics-Constrained Comprehensive Optical Neural Networks

引用

38th Conference on neural Information processing Systems, NeurIPS 2024

作者： Liu, Yanbing Qin, Jianwei Liu, Yan Yue, Xi Liu, Xun Wang, Guoqing Li, Tianyu Ye, Fangwei Li, Wei Beijing University of Posts and Telecommunications Beijing100876 China Shanghai Jiao Tong University Shanghai200240 China Beijing Institute of Space Mechanics and Electricity Beijing100094 China University of Electronic Science and Technology of China Chengdu611731 China

With the advantages of low latency, low power consumption, and high parallelism, optical neural networks (ONN) offer a promising solution for time-sensitive and resource-limited artificial intelligence applications. However, the performance of the ONN model is often diminished by the gap between the ideal simulated system and the actual physical system. To bridge the gap, this work conducts extensive experiments to investigate systematic errors in the optical physical system within the context of image classification tasks. Through our investigation, two quantifiable errors-light source instability and exposure time mismatches-significantly impact the prediction performance of ONN. To address these systematic errors, a physics-constrained ONN learning framework is constructed, including a well-designed loss function to mitigate the effect of light fluctuations, a CCD adjustment strategy to alleviate the effects of exposure time mismatches and a'physics-prior-based' error compensation network to manage other systematic errors, ensuring consistent light intensity across experimental results and simulations. In our experiments, the proposed method achieved a test classification accuracy of 96.5% on the MNIST dataset, a substantial improvement over the 61.6% achieved with the original ONN. For the more challenging QuickDraw16 and Fashion MNIST datasets, experimental accuracy improved from 63.0% to 85.7% and from 56.2% to 77.5%, respectively. Moreover, the comparison results further demonstrate the effectiveness of the proposed physics-constrained ONN learning framework over state-of-the-art ONN approaches. This lays the groundwork for more robust and precise optical computing applications. © 2024 neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

MRI Denoising with Residual Connections and Two-Way Scaling Using Unsupervised Swin Convolutional U-Net Transformer (USCUNT)

MRI Denoising with Residual Connections and Two-Way Scaling ...

引用

International Conference on Applied Soft Computing and Communication networks, ACN 2023

作者： Anu, P. Ramani, G. Hariharasitaraman, S. Robert Singh, A. Athisayamani, Suganya School of Computing Sastra Deemed to be University Tamil Nadu Thanjavur India Department of Computer Applications Sri Sarada College for Women Tamil Nadu Tirunelveli India School of Computing Science and Artificial Intelligence VIT Bhopal University Kothrikalan Madhya Pradesh Sehore India Department of Computational Intelligence SRM Institute of Science and Technology Tamil Nadu Kattankulathur India

ISBN: (纸本)9789819720033

In medical image processing, denoising is a crucial preprocessing step that lowers noise. This paper employs the combination of U-Net and Swin transformer for MRI diagnosis. We combine a convolutional neural network with these two models. Three modules comprise the USCUNT model: 1. Convolution layers are used for extracting spatial features, and 2. Swin transformer is used for image reconstruction. 3. Using UNet for high-level deep feature extraction. In this research, denoising is performed on the artificial noisy MRI pictures. Up and down sampling are used in the UNet module to retrieve deep features. Modern denoising models are contrasted with the suggested architecture. The suggested approach performs better, with an SSIM of 0.823 and PSNR of 29.45. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： image reconstruction

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：