检索结果-内蒙古大学图书馆

Memristive Circuit Design of Quantized Convolutional Auto-Encoder

IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE 2023年第4期7卷 1301-1313页

作者： Zhang, Yaozhong Wang, Xiaoping Yang, Chao Zeng, Zhigang Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automation Wuhan 430074 Peoples R China Huazhong Univ Sci & Technol Key Lab Image Proc & Intelligent Control Educ Minist China Wuhan 430074 Peoples R China Guangxi Univ Sci & Technol Sch Comp Sci & Technol Liuzhou 545006 Peoples R China

This work proposes a memristor-based quantized convolutional auto-encoder (MQCAE) and applies it in an image denoising application. The pulse width or amplitude is gradually tuned by incremental steps in current memristive programming methods, which can be extremely time-consuming. In this work, a multi-level programming method without incremental steps is proposed. MQCAE is composed of five kinds of network layers: convolution layers, deconvolution layers, activation function layers, batch normalization layers and max-pooling layers. We design a memristive circuit for realizing convolution and deconvolution with different kernel parameters. The proposed circuit generates one output feature row every cycle. Analog data buffers are designed to store the intermediate data among network layers. In addition, a reconfigurable analog circuit for realizing activation functions and batch normalization is presented. By using analog temp modules, convolution layers and max-pooling layers are able to compute simultaneously. We construct the MQCAE and introduce the pipeline technique among network layers based on the circuit modules. As a result, MQCAE is able to process one FashionMNIST image in 36 cycles with clock frequency 20 kHz. Finally, we verify the effectiveness of MQCAE in an image denoising task. The results show that the denoising performance of proposed scheme is close to software model while the processing speed is faster.

关键词： Acceleration memristor crossbar array convolutional auto-encoder image denoising quantization

来源：评论

学校读者我要写书评

暂无评论

SU²GE-Net: a saliency-based approach for non-specific class foreground segmentation

引用

SCIENTIFIC REPORTS 2023年第1期13卷 1-11页

作者： Lei, Xiaochun Cai, Xiang Lu, Linjun Cui, Zihang Jiang, Zetao Guilin Univ Elect Technol Sch Comp Sci & Informat Secur Guilin 541010 Guangxi Peoples R China Guilin Univ Elect Technol Guangxi Key Lab Image & Graph Intelligent Proc Guilin 541004 Guangxi Peoples R China

Salient object detection is vital for non-specific class subject segmentation in computer vision applications. However, accurately segmenting foreground subjects with complex backgrounds and intricate boundaries remains a challenge for existing methods. To address these limitations, our study proposes SU(2)GE- Net, which introduces several novel improvements. We replace the traditional CNN-based backbone with the transformer-based Swin-TransformerV2, known for its effectiveness in capturing long-range dependencies and rich contextual information. To tackle under and overattention phenomena, we introduce Gated Channel Transformation (GCT). Furthermore, we adopted an edge-based loss (Edge Loss) for network training to capture spatial-wise structural details. Additionally, we propose Training-only Augmentation Loss (TTA Loss) to enhance spatial stability using augmented data. Our method is evaluated using six common datasets, achieving an impressive F beta score of 0.883 on DUTS-TE. Compared with other models, SU(2)GE- Net demonstrates excellent performance in various segmentation scenarios.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Incomplete label error correction and group joint learning for gland segmentation

引用

BIOMEDICAL SIGNAL procESSING AND CONTROL 2024年 89卷

作者： Wang, Huadeng Liu, Ruitao Zeng, Lingqi Zuo, Wenbin Li, Bingbing Pan, Xipeng Lan, Rushi Luo, Xiaonan Guilin Univ Elect Technol Sch Comp Sci & Informat Secur Guilin 541004 Peoples R China Guangxi Key Lab Image & Graph Intelligent Proc Guilin 541004 Peoples R China Ganzhou Municipal Hosp Dept Pathol Ganzhou 341000 Peoples R China

Medical image segmentation methods based on deep learning show good performance advantages and application prospects, but such methods require experienced medical experts to spend a lot of time labeling data. To address this problem in gland segmentation research, a model named ILECGJL based on incomplete label error correction and group joint learning is proposed, which significantly reduces the dependence on labeled data while ensuring segmentation performance. ILECGJL employs the proposed incomplete label error correction method to capture semantic relationships between labeled and unlabeled images, enhancing the quality of pseudo labels and improving the accuracy of gland segmentation. At the same time, ILECGJL harnesses the proposed group joint learning strategy to enhance the model's robustness on pseudo label noise, further boosting segmentation performance. Experimental results on the CRAG dataset confirm that our proposed label error correction method yields substantial improvements based on initial segmentation predictions. The performance of the ILECGJL model demonstrates significant enhancements compared to fully supervised methods that solely rely on labeled data and outperforms other mainstream semi-supervised approaches. Results and comparisons on the GlaS dataset showcase good generalization ability of the ILECGJL model. Moreover, due to its modular design and customizability, ILECGJL offers high flexibility and can be conveniently integrated into different segmentation models.

关键词： Gland segmentation label error correction Semi-supervised learning Group joint learning Pseudo label

来源：评论

学校读者我要写书评

暂无评论

Spatial-Frequency Discriminability for Revealing Adversarial Perturbations

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 2024年第12期34卷 12608-12623页

作者： Wang, Chao Qi, Shuren Huang, Zhiqiu Zhang, Yushu Lan, Rushi Cao, Xiaochun Fan, Feng-Lei Nanjing Univ Aeronaut & Astronaut Coll Comp Sci & Technol Nanjing 211106 Peoples R China Chinese Univ Hong Kong Dept Math Hong Kong Peoples R China Guilin Univ Elect Technol Guangxi Key Lab Image & G Intelligent Proc Guilin 541004 Peoples R China Sun Yat Sen Univ Sch Cyber Sci & Technol Shenzhen Campus Shenzhen 518107 Peoples R China

The vulnerability of deep neural networks to adversarial perturbations has been widely perceived in the computer vision community. From a security perspective, it poses a critical risk for modern vision systems, e.g., the popular Deep Learning as a Service (DLaaS) frameworks. For protecting deep models while not modifying them, current algorithms typically detect adversarial patterns through discriminative decomposition for natural and adversarial data. However, these decompositions are either biased towards frequency resolution or spatial resolution, thus failing to capture adversarial patterns comprehensively. Also, when the detector relies on few fixed features, it is practical for an adversary to fool the model while evading the detector (i.e., defense-aware attack). Motivated by such facts, we propose a discriminative detector relying on a spatial-frequency Krawtchouk decomposition. It expands the above works from two aspects: 1) the introduced Krawtchouk basis provides better spatial-frequency discriminability, capturing the differences between natural and adversarial data comprehensively in both spatial and frequency distributions, w.r.t. the common trigonometric or wavelet basis;2) the extensive features formed by the Krawtchouk decomposition allows for adaptive feature selection and secrecy mechanism, significantly increasing the difficulty of the defense-aware attack, w.r.t. the detector with few fixed features. Theoretical and numerical analyses demonstrate the uniqueness and usefulness of our detector, exhibiting competitive scores on several deep models and image sets against a variety of adversarial attacks.

关键词： Detectors Perturbation methods Feature extraction Accuracy Training Robustness Artificial neural networks Adversarial example detection orthogonal decomposition spatial-frequency

来源：评论

学校读者我要写书评

暂无评论

Selecting the Best Part From Multiple Laplacian Autoencoders for Multi-View Subspace Clustering

引用

IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING 2023年第7期35卷 7457-7469页

作者： Tang, Kewei Xu, Kaiqiang Jiang, Wei Su, Zhixun Sun, Xiyan Luo, Xiaonan Liaoning Normal Univ Sch Math Dalian 116029 Peoples R China Dalian Univ Technol Sch Math Sci Dalian 116024 Peoples R China Wenzhou Univ Coll Comp Sci & Artificial Intelligence Wenzhou 325035 Peoples R China Dalian Univ Technol Key Lab Computat Math & Data Intelligence Liaoning Dalian 116024 Peoples R China Guilin Univ Elect Technol Guilin 541004 Peoples R China Guilin Univ Elect Technol Guangxi Key Lab Precis Nav Technol & Applicat Guilin 541004 Peoples R China Guilin Univ Elect Technol Sch Informat & Commun Guilin 541004 Peoples R China Guilin Univ Elect Technol Guangxi Key Lab Image & Graph Intelligent Proc Guilin 541004 Peoples R China

The multi-view subspace clustering attracts much attention in recent years. Most methods follow the framework of fusing the affinity graph learned in each view. In this framework, both the fusion strategy and built graph of each view are very important. In this paper, we propose novel methods for multi-view subspace clustering to address these two aspects. On the one hand, we adopt the autoencoders with Laplacian regularization to construct the affinity graph in each view. Compared with previous work employing the autoencoders, the Laplacian term in our method can guide the learned latent representation favoring affinity extraction. Besides, we also discuss the reasons for adding Laplacian regularization. On the other hand, we propose a novel fusion strategy distinguished from the related literature. If the affinity graph of some view is not extracted well, the performance of previous fusion strategies will be seriously affected. Since our strategy can choose the best part from each affinity graph, it can overcome this limitation to some extent. Extensive experimental results on multiple benchmark data sets confirm the effectiveness of our method.

关键词： Laplace equations Neural networks Deep learning Faces Clustering methods Sun Feature extraction Multi-view subspace clustering autoencoders Laplacian operator novel fusion strategy

来源：评论

学校读者我要写书评

暂无评论

A Skeleton Posture Transfer Method from Kinect Capture

A Skeleton Posture Transfer Method from Kinect Capture

引用

International Conference on Virtual Reality, Human-Computer Interaction and Artificial Intelligence (VRHCIAI)

作者： Deng, Zhengjie He, Shuctian Zhang, Mengyan Wang, Yizhen Hainan Nounal Univ Sch Informat Sci & Technol Haikou Peoples R China Guilin Univ Elect Technol Guangxi Key Lab Image & Graph Intelligent Proc Guilin Peoples R China Hainan Noimal Univ Coll Phys & Elect Engn Haikou Peoples R China

ISBN: (纸本)9781665491822

At present, 3D models are applied to many fields, and the action control of 3D model needs be studied further. In order to control a 3D model using user-acceptable devices, this paper proposes a method to transfer the posture of the skeleton captured by Kinect to a skeleton of the Soldier model, which is a sample in Microsoft's DirectX SDK First, using the Kinect to collect the user's skeleton information, combined with the initial posture of the Soldier skeleton, through node matching, inverse mirror transformation, transformation from absolute to relative, coordinate space transformation, to generate a skeleton with the same structure as the Soldier skeleton, which can be applied to the Soldier model. And in a virtual reality scene, the corresponding actions of users can drive the Soldier model. Experiments show that the method is feasible and effective. The proposed method can also be referenced and applied to a target skeleton with other structures.

关键词： posture transfer skeleton Kinect 3D model virtual reality

来源：评论

学校读者我要写书评

暂无评论

Full-Circuit Implementation of Transformer Network Based on Memristor

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS 2022年第4期69卷 1395-1407页

作者： Yang, Chao Wang, Xiaoping Zeng, Zhigang Huazhong Univ Sci & Technol Inst Artificial Intelligence Sch Artificial Intelligence & Automat Wuhan 430074 Peoples R China Huazhong Univ Sci & Technol Key Lab Image Proc & Intelligent Control Educ Minist China Wuhan 430074 Peoples R China Guangxi Univ Sci & Technol Sch Elect Elect & Comp Sci Liuzhou 545006 Peoples R China Huazhong Univ Sci & Technol Key Lab Image Proc & Intelligent Control Educ Minist China Sch Artificial Intelligence & Automat Wuhan 430074 Peoples R China

As an emerging in-memory element, memristor has been widely used in various neural network circuits to represent the weights and accelerate the calculation. However, the Transformer Network (TN), one of the most important models for machine vision and natural language processing in recent years, has not yet been full-circuit implemented using memristors due to the complex calculation process and data storage. In order to carry out the computation of the TN more efficiently, this work proposes a memristor-based full-circuit implementation of the TN capable of: 1) a memristor crossbar module to preserve the weights of the TN and perform the vector-matrix multiplications;2) an analog signal memory module to store the analog signal directly in near-memory mode;3) function circuit modules to achieve five transformations, namely Softmax, Layer Normalization, ReLU, Multiply-add and Residual;4) a timing signal generation module to schedule operations of the circuit. The proposed TN circuit can complete all calculations directly based on the analog signal without using any analog-digital converter (ADC), digital-analog converter (DAC) and digital memory. In addition, character image recognition experiments are carried out in PSPICE to verify the functional correctness of the designed circuit. The corresponding signal retention rates of the analog memory, the performance of the whole circuit, and the non-idealities of the memristors are also analyzed. The results indicate that the circuit has advantages in terms of area overhead, energy efficiency and anti-noise.

关键词： Memristors Transformers Neural networks Data models Task analysis image recognition Computational modeling Memristor transformer network analog signal memory circuit implementation

来源：评论

学校读者我要写书评

暂无评论

Simple Conditional Spatial Query Mask Deformable Detection Transformer: A Detection Approach for Multi-Style Strokes of Chinese Characters

引用

SENSORS 2024年第3期24卷 931页

作者： Zhou, Tian Xie, Wu Zhang, Huimin Fan, Yong Guilin Univ Elect Technol Guangxi Key Lab Image & Graph Intelligent Proc Guilin 541004 Peoples R China Guilin Univ Elect Technol Sch Comp Sci & Informat Secur Guilin 541004 Peoples R China Guangxi Normal Univ Minist Educ Key Lab Educ Blockchain & Intelligent Technol Guilin 541004 Peoples R China Guangxi Normal Univ Guangxi Key Lab Multisource Informat Min & Secur Guilin 541004 Peoples R China Guilin Univ Elect Technol Sch Mech & Elect Engn Guilin 541004 Peoples R China

In the Chinese character writing task performed by robotic arms, the stroke category and position information should be extracted through object detection. Detection algorithms based on predefined anchor frames have difficulty resolving the differences among the many different styles of Chinese character strokes. Deformable detection transformer (deformable DETR) algorithms without predefined anchor frames result in some invalid sampling points with no contribution to the feature update of the current reference point due to the random sampling of sampling points in the deformable attention module. These processes cause a reduction in the speed of the vector learning stroke features in the detection head. In view of this problem, a new detection method for multi-style strokes of Chinese characters, called the simple conditional spatial query mask deformable DETR (SCSQ-MDD), is proposed in this paper. Firstly, a mask prediction layer is jointly determined using the shallow feature map of the Chinese character image and the query vector of the transformer encoder, which is used to filter the points with actual contributions and resample the points without contributions to address the randomness of the correlation calculation among the reference points. Secondly, by separating the content query and spatial query of the transformer decoder, the dependence of the prediction task on the content embedding is relaxed. Finally, the detection model without predefined anchor frames based on the SCSQ-MDD is constructed. Experiments are conducted using a multi-style Chinese character stroke dataset to evaluate the performance of the SCSQ-MDD. The mean average precision (mAP) value is improved by 3.8% and the mean average recall (mAR) value is improved by 1.1% compared with the deformable DETR in the testing stage, illustrating the effectiveness of the proposed method.

关键词： object detection Chinese character stroke transformer deformable DETR SCSQ-MDD

来源：评论

学校读者我要写书评

暂无评论

H8 Controller Design for Networked Systems With Two-Channel Packet Dropouts and FDI Attacks

引用

IEEE TRANSACTIONS ON CYBERNETICS 2024年第3期54卷 1661-1670页

作者： Hu, Zhipei Chen, Kai Deng, Feiqi Luo, Shixian Hu, Songlin Shantou Univ Coll Engn Dept Elect & Informat Engn Guangdong Prov Key Lab Digital Signal & Image Proc Shantou 515063 Peoples R China Shantou Univ Key Lab Intelligent Mfg Technol Minist Educ Shantou 515063 Peoples R China South China Univ Technol Coll Automat Sci & Engn Guangzhou 510640 Peoples R China Guangxi Univ Sch Elect Engn Nanning 530004 Peoples R China Nanjing Univ Posts & Telecommun Inst Adv Technol Nanjing 210023 Peoples R China

In this article, the stochastic analysis and Hop controller design problems of networked systems with packet dropouts and false data injection attacks are investigated. Different from the existing literature, we focus on the linear networked systems with external disturbances and both sensor- controller channel and controller-actuator channel are studied. First, we present a discrete-time modeling framework that leads to a stochastic closed-loop system with randomly varying parameters. To facilitate the analysis and H-infinity control of resulting discrete-time stochastic closed-loop system, an equivalent yet analyzable stochastic augmented model is further constructed by matrix exponential computation. Based on this model, a stability condition is derived in the form of linear matrix inequality (LMI) with the aid of a reduced-order confluent Vandermonde matrix, Kronecker product operation, and law of total expectation. Specifically, the dimension of the LMI obtained in this article does not increase as the upper bound of consecutive packet dropouts does, which is also different from the existing literature. Subsequently, a desired H-infinity controller is obtained such that the original discrete-time stochastic closed-loop system is exponentially mean-square stable with a prescribed H-infinity performance. Finally, a numerical example and a direct current motor system are exploited to substantiate the effectiveness and practicability of the designed strategy.

关键词： Stochastic processes Linear matrix inequalities Cyberattack Computational modeling Stability analysis Data models DC motors Discrete-time approach false data injection (FDI) attacks networked control systems packet dropouts sampled-data systems

来源：评论

学校读者我要写书评

暂无评论

Usability Enhanced Thumbnail-Preserving Encryption Based on Data Hiding for JPEG images

引用

IEEE SIGNAL procESSING LETTERS 2023年 30卷 793-797页

作者： Ye, Xi Zhang, Yushu Xiao, Xiangli Yi, Shuang Lan, Rushi Nanjing Univ Aeronaut & Astronaut Coll Comp Sci & Technol Nanjing 211106 Peoples R China Chinese Acad Sci Inst Informat Engn State Key Lab Informat Secur Beijing 100093 Peoples R China Southwest Univ Polit Sci & Law Criminal Invest Sch Chongqing 401120 Peoples R China Guilin Univ Elect Technol Guangxi Key Lab Image & Grap Intelligent Proc Guilin 541004 Peoples R China

As an increasing number of people tend to upload personal images to cloud platforms such as iCloud, Google Drive, and Baidu Cloud, the issue of image privacy protection in clouds has attracted widespread attention. Thumbnail-preserving encryption (TPE) is a newly proposed technology especially for balancing usability and privacy of images stored in clouds. The TPE-encrypted image does not reveal anything other than the thumbnail and size of the original one. Although many well-designed TPE schemes have been put forward, they almost all operate in the spatial domain and are not compatible with JPEG images. The only two schemes that can be applied in the frequency domain were proposed by Wright et al. and Marohn et al., but both of them have disadvantages and do not consider non-visual usability. In this letter, we propose a usability enhanced TPE scheme especially for JPEG images by utilizing reversible data hiding and image encryption in the frequency domain. We first transform the original image into the frequency domain and select some specific locations for extra information recording in every block. The original coefficient values at the selected locations are then collected and embedded into the rest of the block before encryption. Finally, the extra information is expected to be recorded into the selected locations of every block for non-visual usability. Experiments confirm that the proposed scheme achieves the target of balancing enhanced usability and privacy.

关键词： Reversible data hiding thumbnail-preserving encryption visual privacy protection

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：