检索结果-内蒙古大学图书馆

Simultaneously Learning Deep Quaternion Reconstruction and Noise Convolutional Dictionary for Color image Denoising

ieee TRANSACTIONS ON EMERGING TOPICS IN computational intelligence 2025年第2期9卷 1766-1779页

作者： Zhou, Zheng Chen, Yongyong Zhou, Yicong Univ Macau Dept Comp & Informat Sci Macau 999078 Peoples R China Harbin Inst Technol Sch Comp Sci & Technol Shenzhen 518055 Peoples R China

Recently, many deep convolutional dictionary learning-based methods, integrating the traditional image representation methods with deep neural networks, have achieved great success in various image processing tasks. However, the existing approaches can be further improved with the following considerations: (1) They congenitally suffer from the high cross-channel correlation loss for color image processing tasks since they usually treat each color channel independently, not in a whole perspective. (2) They only build up a single reconstruction dictionary learning model to directly approximate images using several single dictionary atoms, which cannot make full use of the representative ability of the model. In this paper, we propose a simultaneously learning deep quaternion reconstruction and noise convolutional dictionary model. To fully explore the cross-channel correlation, we use the quaternion method to process the color image in a holistic way. An adaptive attentional weight of reconstruction and noise learning module is also developed for the optimal combination between reconstruction and noise learning. Experimental results for synthesis and real color image denoising have demonstrated the superiority of the proposed method over other state-of-the-art methods.

关键词： Quaternions Dictionaries Color image reconstruction image color analysis Convolution Noise reduction Deep unfolding learning color image denoising quaternion

来源：评论

学校读者我要写书评

暂无评论

EdgeActNet: Edge intelligence-Enabled Human Activity Recognition Using Radar Point Cloud

引用

ieee TRANSACTIONS ON MOBILE COMPUTING 2024年第5期23卷 5479-5493页

作者： Luo, Fei Khan, Salabat Li, Anna Huang, Yandao Wu, Kaishun Shenzhen Univ Coll Comp Sci & Software Engn Shenzhen 518060 Peoples R China Queen Mary Univ London Sch Elect Engn & Comp Sci London E1 4NS England Hong Kong Univ Sci & Technol Guangzhou 510230 Peoples R China

Human activity recognition (HAR) has become a research hotspot because of its wide range of application prospects. It has higher requirements for real-time and power-efficient processing. However, a large amount of data transfer between sensors and servers, and computation-intensive recognition models hinder the implementation of real-time HAR systems. Recently, edge computing has been proposed to address this challenge by moving computational and data storage resources to the sensors, rather than depending on a centralized server/cloud. In this paper, we investigated binary neural networks for edge intelligence-enabled HAR using radar point cloud. Point cloud can provide 3-dimensional spatial information, which is helpful to improve recognition accuracy. Time-series point cloud also brings challenges, such as larger data volume, 4-dimensional data processing, and more intensive computation. To tackle these challenges, we adopt the 2-dimensional histograms for point cloud multi-view processing and propose the EdgeActNet, a binary neural network for point cloud-based human activity classification on edge devices. In the evaluation, the EdgeActNet achieved the best results with average accuracies of 97.63% on the MMActivity dataset and 95.03% on the point cloud samples of the DGUHA dataset respectively;and saved $16.9\times$16.9x memory consumption and 11.5x inference time compared to its full-precision version. Our work also is the first to apply 2D histogram-based multi-view representation and BNNs for time-series point cloud classification.

关键词： Point cloud compression Human activity recognition Radar Sensors image edge detection computational modeling Real-time systems radar point cloud binary neural network edge intelligence

来源：评论

学校读者我要写书评

暂无评论

Multilogic Sense Amplifier With a Circuit for Dynamic Reference Voltage Generation

引用

ieee TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS 2025年

作者： Chang, Yen-Jen Tsai, Kun-Lin Cheng, Chun Chen, Xu-Yao Natl Chung Hsing Univ Dept Comp Sci & Engn Taichung 402202 Taiwan Tunghai Univ Dept Elect Engn Taichung 407224 Taiwan Tunghai Univ Res Ctr Smart Sustainable Circular Econ Taichung 407224 Taiwan

The rapid development of artificial intelligence (AI) systems has engendered a considerable increase in the computational power required in data-intensive applications, including facial recognition and image processing applications. The conventional von Neumann architecture, in which large quantities of data are transmitted between processing units and memory units, creates a bottleneck in AI applications. In-memory computing (IMC) offers an efficient solution to this problem by enabling computations within memory units, thus reducing the need for data transmission. This paper proposes a multilogic sense amplifier (MSA) with a dynamic reference voltage (DRV) generation circuit (hereafter denoted as MSA-DRV) to enhance the performance and reduce the power consumption of static random-access memory (SRAM)-based IMC. The proposed MSA-DRV performs six logic operations, namely the AND, NAND, OR, NOR, XOR, and XNOR operations, within an SRAM array by using a novel DRV circuit. The DRV circuit enables the voltage threshold to be adaptively changed according to the requirements of different operations. Experimental results indicated that the proposed MSA-DRV had lower computational latency (on average 29.9% lower) and power consumption (on average 30.6% lower) than did a conventional sense amplifier design and reconfigurable assist sense amplifier. Thus, the proposed design can overcome the von Neumann bottleneck to facilitate high-speed, energy-efficient data processing, which is crucial for AI-based and other data-intensive applications.

关键词： Circuits Logic Transistors Discharges (electric) Power demand SRAM cells Artificial intelligence Voltage Logic gates High-voltage techniques Dynamic reference voltage (DRV) in-memory computing (IMC) multilogic operations sense amplifier

来源：评论

学校读者我要写书评

暂无评论

Optical Illusions Recognition intelligence 8

Optical Illusions Recognition Intelligence

引用

8th ieee symposium on Wireless Technology and Applications, ISWTA 2024

作者： Leong, Wai Yie Leong, Yuan Zhi Leong, Wai San INTI International University Persiaran Perdana BBN Putra Nilai Negeri Sembilan Nilai71800 Malaysia Schneider Electric Singapore 50 Kallang Avenue Kallang Singapore Singapore

ISBN: (纸本)9798350351354

Optical illusions have intrigued humanity for centuries, challenging our perception and understanding of visual stimuli. This paper delves into the realm of optical illusions through the lens of image processing techniques. By employing various digital tools and algorithms, we aim to dissect and analyze the mechanisms behind these illusions, shedding light on the fascinating interplay between human vision and computational models. Through a series of experiments and demonstrations, we showcase how image processing not only elucidates the inner workings of optical illusions but also unveils the potential applications of such insights in diverse fields, ranging from psychology to computer vision. © 2024 ieee.

关键词： Digital elevation model

来源：评论

学校读者我要写书评

暂无评论

Robust Model Watermarking for image processing Networks via Structure Consistency

引用

ieee TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE intelligence 2024年第10期46卷 6985-6992页

作者： Zhang, Jie Chen, Dongdong Liao, Jing Ma, Zehua Fang, Han Zhang, Weiming Feng, Huamin Hua, Gang Yu, Nenghai Univ Sci & Technol China Sch Cyber Sci & Secur Hefei 230026 Anhui Peoples R China Microsoft Res Redmond WA 98052 USA City Univ Hong Dept Comp Sci Kowloon Tong Hong Kong Peoples R China Beijing Elect Sci & Technol Inst Beijing 100070 Peoples R China Wormpex AI Res LLC Bellevue WA 98004 USA

The intellectual property of deep networks can be easily "stolen" by surrogate model attack. There has been significant progress in protecting the model IP in classification tasks. However, little attention has been devoted to the protection of image processing models. By utilizing consistent invisible spatial watermarks, the work (Zhang et al. 2020) first considered model watermarking for deep image processing networks and demonstrated its efficacy in many downstream tasks. Its success depends on the hypothesis that if a consistent watermark exists in all prediction outputs, that watermark will be learned into the attacker's surrogate model. However, when the attacker uses common data augmentation attacks (e.g., rotate, crop, and resize) during surrogate model training, it will fail because the underlying watermark consistency is destroyed. To mitigate this issue, we propose a new watermarking methodology, "structure consistency", based on which a new deep structure-aligned model watermarking algorithm is designed. Specifically, the embedded watermarks are designed to be aligned with physically consistent image structures, such as edges or semantic regions. Experiments demonstrate that our method is more robust than the baseline in resisting data augmentation attacks. Besides that, we test the generalization ability and robustness of our method to a broader range of adaptive attacks.

关键词： Watermarking Data augmentation image processing Task analysis Training computational modeling Data models Deep model IP protection model watermarking image processing

来源：评论

学校读者我要写书评

暂无评论

Fundus image Enhancement via Semi-Supervised GAN and Anatomical Structure Preservation

引用

ieee TRANSACTIONS ON EMERGING TOPICS IN computational intelligence 2024年第1期8卷 313-326页

作者： Wu, Hao-Tian Cao, Xin Gao, Ying Zheng, Kaihan Huang, Jiwu Hu, Jiankun Tian, Zhihong South China Univ Technol Sch Comp Sci & Engn Guangzhou 510006 Peoples R China Guangdong Prov Key Lab Artificial Intelligence Med Guangzhou Peoples R China Shenzhen Univ Guangdong Key Lab Intelligent Informat Proc Shenzhen 518060 Peoples R China Shenzhen Univ Shenzhen Key Lab Media Secur Shenzhen 518060 Peoples R China Shenzhen Inst Artificial Intelligence & Robot Soc Shenzhen Peoples R China Univ New South Wales Australian Def Force Acad Sch Engn & Informat Technol Canberra ACT 2612 Australia Guangzhou Univ Cyberspace Inst Adv Technol Guangzhou 511370 Peoples R China

The fundus image is often used in clinical diagnosis due to ease and safety of acquisition, but the quality may be affected by environment and onsite operations. As low-quality medical images may lead to misinterpretation in diagnosis and analysis, it is important to improve quality of the improperly acquired fundus images. Unfortunately, the existing fundus image enhancement methods require task-specific prior knowledge or suffer from insufficient generalization ability. To cope with this issue, a generative adversarial network (GAN) based model is proposed, namely the semi-supervised GAN with anatomical structure preservation (SSGAN-ASP). Specifically, an anatomical structure extraction component is employed in the generator to guide the enhancement process by preserving both retinal and lesion structures, while color information in the fundus image is also preserved. The SSGAN-ASP model is evaluated and compared with the state-of-the-art methods for medical image enhancement on three popular datasets. In addition, it is applied in the pre-processing of retinal vessel segmentation and diabetic retinopathy grading tasks to show efficacy in computer-aided diagnosis. Experimental results demonstrate that visual quality of the enhanced image can be improved while better performance in clinical diagnosis is achieved with our proposed model by adopting the anatomical structure extraction component and preserving color information as well.

关键词： Fundus image enhancement anatomical structure generative adversarial network semi-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Energy-Efficient Deep Learning for Cloud Detection Onboard Nanosatellite

引用

ieee JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2025年 18卷 9968-9985页

作者： Khalil, Imane Daghouri, Amina Chanoui, Mohammed Alae Guennoun, Zouhair Addaim, Adnane Mohammed V Univ Univ Ctr Res Space Technol Mohammadia Sch Engineers Rabat Morocco

Deep learning has been increasingly utilized for cloud detection in Earth observation nanosatellites, offering effective solutions to enhance mission performance. Traditional methods of image capture, onboard storage, and transmission face challenges such as bandwidth limitations and the transmission of cloud-obstructed images, highlighting the need for efficient onboard artificial intelligence. However, energy consumption remains a critical consideration for onboard processing, particularly in nanosatellites where resource constraints are significant. To address these challenges, we propose an optimized SegNet-based deep learning model implemented on the STM32H7 microcontroller for real-time cloud detection, designed to operate within the nanosatellite's strict energy budget. This work, conducted within the project of the UM5-EOSat nanosatellite mission, utilized captured images from the Gecko imager for model evaluation. The customized SegNet architecture, tailored with minimal kernels and layers, achieved an accuracy of 93.50%, effectively balancing performance and computational efficiency. Quantization further optimized energy consumption, achieving a reduction of 82.2% at 280 MHz. The quantized model demonstrated a memory footprint of 304 KB RAM and 110 KB Flash memory, with an inference time of 0.21 s and an energy consumption of 31.41 mJ, ensuring compatibility with the nanosatellite's resource constraints.

关键词： computational modeling Accuracy Energy consumption Clouds Memory management Artificial intelligence Deep learning computational efficiency Power demand Microcontrollers Cloud detection deep learning energy consumption nanosatellite TinyML

来源：评论

学校读者我要写书评

暂无评论

A Novel Clutter Suppression Algorithm for Airborne Radar via Subspace Projection

引用

ieee SENSORS JOURNAL 2025年第1期25卷 962-970页

作者： Zhang, Xinying Wang, Tong Cui, Weichen Wang, Degen Xidian Univ Natl Key Lab Radar Signal Proc Xian 710071 Peoples R China Beijing Aerosp Automatic Control Inst Natl Key Lab Sci & Technol Aerosp Intelligence Con Beijing 100854 Peoples R China Southwest China Res Inst Elect Equipment Chengdu 610036 Peoples R China

Strong ground clutter submerges the moving targets and hinders the radar image analysis. An effective technique called space-time adaptive processing (STAP) is used to suppress clutter in airborne radar systems but has the disadvantage of large training sample requirements. Sparse recovery STAP (SR-STAP) has the benefit of improving the algorithm performance when the training sample size is small. However, the computational burden and storage space restrict the practical application of SR-STAP. To address this issue, a fast subspace-based SR-STAP algorithm is proposed in this article. Using the Bayesian framework, the proposed algorithm selects the most matched atoms to span a clutter subspace by minimizing the likelihood function. In comparison with the existing SR-STAP algorithm, the proposed algorithm avoids the time-consuming sparse reconstruction process. Moreover, the estimated subspace can be applied to gain-phase error calibration. The proposed algorithm is experimentally demonstrated to have superior performances in terms of clutter suppression, target detection, and computational complexity.

关键词： Clutter Vectors Noise Training Sensors Airborne radar Signal processing algorithms Object detection Bayes methods Atoms Clutter suppression gain-phase errors space-time adaptive processing (STAP) sparse recovery subspace estimated

来源：评论

学校读者我要写书评

暂无评论

Classification of Cochannel Signals using Cyclostationary Signal processing and Deep Learning

Classification of Cochannel Signals using Cyclostationary Si...

引用

ieee International symposium on Dynamic Spectrum Access Networks (ieee DySPAN)

作者： Mehta, Tanay Crompton, Bryan Mody, Apurva Northeastern Univ Dept Comp Sci Boston MA 02115 USA AiRANACULUS Lowell MA USA

ISBN: (纸本)9798350317657;9798350317640

The task of classifying radio frequency (RF) signals has seen recent success in employing deep neural network models. In this work, we present a combined signal processing and machine learning approach to signal classification for cochannel anomalous signals. The power spectral density and cyclostationary signal processing features of a captured signal are computed and fed into a neural net to produce a classification decision. Experiments show that our neural model utilizing both PSD and CSP features are superior to models only using CSP features, and achieve near perfect classification. Our combined signal preprocessing and machine learning approach allows for simpler neural networks with fast training times and small computational resource requirements for inference with longer preprocessing time.

关键词： signal classification machine learning deep learning cochannel interference

来源：评论

学校读者我要写书评

暂无评论

Guest Editorial Special Issue on Explainable Deep Learning for Medical image processing and Analysis

引用

ieee TRANSACTIONS ON EMERGING TOPICS IN computational intelligence 2023年第1期7卷 2-4页

作者： Zhang, Yu-Dong Gorriz, Juan Manuel Pan, Yi Cordon, Oscar Univ Leicester Leicester LE1 7RH England Univ Cambridge Cambridge CB2 1TN England Chinese Acad Sci Shenzhen Inst Adv Technol Shenzhen 518055 Peoples R China Univ Granada Dept Comp Sci Granada 18071 Spain Univ Granada Artificial Intelligence & Andalusian Res Inst Data Granada 18071 Spain

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：