检索结果-内蒙古大学图书馆

Proceedings of the Digital Image Computing: Technqiues and Applications (DICTA)

作者： Inzela Mirza Shahzaib Iqbal Bandar Alhaqbani Bandar AlShammari Tariq Khan Imran Razzak TCC Research and Development Labs Technology Control Company Riyadh Saudi Arabia Department of Computing Abasyn University Islamabad Campus(AUIC) Islamabad Pakistan Computer and Information Sciences College Jouf University Saudi Arabia School of Computer Science and Engineering University of New South Wales Sydney Australia

ISBN: (数字)9798350379037

ISBN: (纸本)9798350379044

Over recent years, deep convolutional neural networks have advanced the field of face recognition in both verification and identification applications. In this paper, we present ShadowNets, a highly efficient and compact face recognition network. We employed multiple shadow blocks in a cascading manner where the output of each shadow block is given as the input to the next shadow block. By integrating the advantages of Focal Modulation Blocks, Shadow Block and Ghost Modules, our proposed ShadowNets achieves superior face recognition performance, optimized specifically for edge devices. This network architecture not only minimizes computational costs and storage needs but also delivers high accuracy, making it suitable for deployment in edge environments. Our extensive experiments on challenging benchmark face datasets showcase the outstanding performance by achieving accuracy 98.85% on LFW, 84.37% on CFP-FP, and 91.23% on AgeDB-30 which validates the effectiveness of ShadowNets over state-of-the-art lightweight and deep face recognition models.

关键词： Training Performance evaluation Accuracy Image recognition Face recognition Computational modeling Modulation Benchmark testing Robustness Computational efficiency

来源：评论

学校读者我要写书评

暂无评论

Feature Extraction of Motor Imagination EEG Signals for a Collaborative Exoskeleton Robot Based on PSD Analysis

Feature Extraction of Motor Imagination EEG Signals for a Co...

引用

Chinese control Conference (CCC)

作者： Yunfeng Wu Junze Li Zichen Ren Bo Lu Changzhu Zhang Peng Qi Department of Control Science and Engineering Tongji University Shanghai China Elmore Family School of Electrical and Computer Engineering Purdue University West Lafayette Indiana United States Robotics and Microsystems Center School of Mechanicaland Electric Engineering Soochow University Suzhou China

The integration of brain-machine interface and exoskeleton robot has been widespread application in gait correction, walking assistance, and numerous other scenarios. To effectively extract the electroencephalogram (EEG) signal features of motor imagery while wearing an exoskeleton, this study proposes a frequency band pre-determination method based on power spectral density (PSD) that enables common spatial patterns (CSP) to extract features from the frequency band with the highest energy in the EEG. The signal power spectral density of all channels is obtained at or near the average frequency of the maximum short interval frequency component energy. A second round of filtering is performed on the data, and the signal components are used as input for the subsequent feature extraction step. The CSP method extracts features from the spatial domain signal and generates feature maps. Finally, Support Vector Machines (SVM) are utilized to classify the EEG signals. Based on the pre-set gait of the exoskeleton robot and the motor imagery paradigm, feature extraction and classification of motor imagery EEG data during exoskeleton use were conducted, with an average accuracy rate of 77%. The experimental results demonstrate the effectiveness of this method in extracting the motor imagery EEG features during exoskeleton use.

关键词：

来源：评论

学校读者我要写书评

暂无评论

EQUILIBRIA AND STABILITY OF ONE MESSENGER RNA AND TWO MICRO RNA DYNAMICS

UPB Scientific Bulletin, Series C: Electrical Engineering an...

引用

UPB Scientific Bulletin, Series C: Electrical engineering and computer science 2022年第4期84卷 3-8页

作者： Olteanu, Mircea Ştefan, Radu Department of Mathematical Methods and Models Faculty of Applied Sciences University "Politehnica" of Bucharest Romania Department of Automatic Control and Systems Engineering Faculty of Automatic Control and Computer Science University "Politehnica" of Bucharest Romania

来源：评论

学校读者我要写书评

暂无评论

Supervised Feature Compression based on Counterfactual Analysis

arXiv

引用

arXiv 2022年

作者： Piccialli, Veronica Morales, Dolores Romero Salvatore, Cecilia Department of Computer Control and Management Engineering Antonio Ruberti Sapienza University of Rome Rome Italy Department of Economics Copenhagen Business School Frederiksberg Denmark Department of Civil Engineering and Computer Science University of Rome Tor Vergata Rome Italy

Counterfactual Explanations are becoming a de-facto standard in post-hoc interpretable machine learning. For a given classifier and an instance classified in an undesired class, its counterfactual explanation corresponds to small perturbations of that instance that allows changing the classification outcome. This work aims to leverage Counterfactual Explanations to detect the important decision boundaries of a pre-trained black-box model. This information is used to build a supervised discretization of the features in the dataset with a tunable granularity. Using the discretized dataset, an optimal Decision Tree can be trained that resembles the black-box model, but that is interpretable and compact. Numerical results on real-world datasets show the effectiveness of the approach in terms of accuracy and sparsity. Copyright © 2022, The Authors. All rights reserved.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

Resource-Efficient and Self-Adaptive Quantum Search in a Quantum-Classical Hybrid System

arXiv

引用

arXiv 2024年

作者： Jiang, Zihao Du, Zefan Ruan, Shaolun Chen, Juntao Wang, Yong Cheng, Long Buyya, Rajkumar Mao, Ying The Department of Computer and Information Science Fordham University New York City United States The School of Computing and Information Systems Singapore Management University Singapore The School of Control and Computer Engineering North China Electric Power University Beijing China The School of Computing and Information Systems The University of Melbourne Australia

Over the past decade, the rapid advancement of deep learning and big data applications has been driven by vast datasets and high-performance computing systems. However, as we approach the physical limits of semiconductor fabrication in the post-Moore's Law era, questions arise about the future of these applications. In parallel, quantum computing has made significant progress with the potential to break limits. Major companies like IBM, Google, and Microsoft provide access to noisy intermediate-scale quantum (NISQ) computers. Despite the theoretical promise of Shor's and Grover's algorithms, practical implementation on current quantum devices faces challenges, such as demanding additional resources and a high number of controlled operations. To tackle these challenges and optimize the utilization of limited onboard qubits, we introduce ReSaQuS, a resource-efficient index-value searching system within a quantum-classical hybrid framework. Building on Grover's algorithm, ReSaQuS employs an automatically managed iterative search approach. This method analyzes problem size, filters less probable data points, and progressively reduces the dataset with decreasing qubit requirements. Implemented using Qiskit and evaluated through extensive experiments, ReSaQuS has demonstrated a substantial reduction, up to 86.36% in cumulative qubit consumption and 72.72% in active periods, reinforcing its potential in optimizing quantum computing application deployment. © 2024, CC BY.

关键词： Hybrid systems

来源：评论

学校读者我要写书评

暂无评论

The subadditivity of quantum entropy in Gaussian quantum systems

引用

Results in Physics 2024年 59卷

作者： Liu, Lijun Ge, Xiaozhen Cheng, Shuming Department of Mathematics and Computer Science Shanxi Normal University Taiyuan 030006 China Shanghai Institute of Intelligent Science and Technology Tongji University Shanghai 201804 China The Department of Control Science and Engineering Tongji University Shanghai 201804 China Institute for Advanced Study Tongji University Shanghai 200092 China

The entropy of a quantum state measures information or uncertainty contained in the quantum system and plays a crucial role in quantum information theory. Many entropic properties, such as the subadditivity, have been studied for the discrete system with a finite-dimensional state space, while much less is explored in the infinite-dimensional system. In this work, we study the entropic properties of the continuous system that models a collection of Gaussian modes or fields. In particular, we identify the parameter range for the Rényi, Tsallis, and Unified entropies of Gaussian states that they are additive, subadditive, and strong subadditive. We show that the Rényi entropies are subadditive, further giving rise to the subadditivity of the Tsallis and Unified entropies in certain parameter range. We also find that the Tsallis and Unified entropies are not strong subadditive, while the Rényi entropies with the order ranging from 1 to 2 are conjectured to admit the strong subadditivity. Our results are useful in studying quantum correlations and channel capacities in Gaussian systems. © 2024 The Author(s)

关键词：

来源：评论

学校读者我要写书评

暂无评论

Towards Trustworthy AI: Sandboxing AI-Based Unverified controllers for Safe and Secure Cyber-Physical Systems

Towards Trustworthy AI: Sandboxing AI-Based Unverified Contr...

引用

IEEE Conference on Decision and control

作者： Bingzhuo Zhong Siyuan Liu Marco Caccamo Majid Zamani TUM School of Engineering and Design Technical University of Munich Germany Division of Decision and Control Systems KTH Royal Institute of Technology Stockholm Sweden Department of Computer Science University of Colorado Boulder USA

In the past decade, artificial-intelligence-based (AI-based) techniques have been widely applied to design controllers over cyber-physical systems (CPSs) for complex control missions (e.g., motion planning in robotics). Nevertheless, AI-based controllers, particularly those developed based on deep neural networks, are typically very complex and are challenging to be formally verified. To cope with this issue, we propose a secure-by-construction architecture, namely Safe-Sec-visor architecture, to sandbox AI-based unverified controllers. By applying this architecture, the overall safety and security of CPSs can be ensured simultaneously, while formal verification over the AI-based controllers is not required. Here, we consider invariance and opacity properties as the desired safety and security properties, respectively. Accordingly, by leveraging a notion of (augmented) control barrier functions, we design a supervisor to check the control inputs provided by the AI-based controller and decide whether to accept them. At the same time, a safety-security advisor runs in parallel and provides fallback control inputs whenever the AI-based controller is rejected for safety and security reasons. To show the effectiveness of our approaches, we apply them to a case study on a quadrotor controlled by an AI-based controller. Here, the initial state of the quadrotor contains secret information which should not be revealed while the safety of the quadrotor should be ensured.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Privacy computing using deep compression learning techniques for neural decoding

引用

Smart Health 2022年 23卷

作者： Li, Huining Chen, Huan Xu, Chenhan Das, Anarghya Chen, Xingyu Li, Zhengxiong Xiao, Jian Huang, Ming-Chun Xu, Wenyao Department of Computer Science and Engineering University at Buffalo United States Department of Electrical Computer and Systems Engineering Case Western Reserve University United States Department of Computer Science and Engineering University of Colorado Denver United States School of Electronic and Control Engineering Chang'an University China Department of Data and Computational Science Duke Kunshan University China

The brain–computer interface supports a variety of applications with the help of machine learning technology. However, existing edge-cloud infrastructure requires subjects to send their sensitive neural signals to the cloud for training the model which brings privacy concerns. Although the recent distributed learning technology is used to help protect subjects’ privacy, it brings high communication costs and cannot avoid privacy reconstruction attacks. In this paper, we propose deep compression learning techniques in the privacy computing infrastructure that can be used for neural decoding while preserving privacy and largely reducing the communication cost. Specifically, we first perform heterogeneous neural signals processing and convert them to resized functional brain connectivity images. Then, a semantics structure-based unsupervised deep compression learning network is trained and generates a neural hash locally for each image. Each hash value is irreversible that cannot be used to reconstruct the user's original neural signal. After that, the cloud end receives the uploaded neural hashes and corresponding labels and filters the abnormal ones. Finally, these neural hashes are used for training neural decoding models. Since a single hash value can correspond to different types of labels, it only needs to be uploaded once with a very small size and then reused for different training tasks, which largely reduce communication cost. Our experiment results show that the proposed privacy computing techniques can be applied to heterogeneous neural signals for training different neural decoding models where the relative accuracy can achieve above 83%. © 2021

关键词： Brain–computer interface Neural decoding Privacy computing Unsupervised deep learning

来源：评论

学校读者我要写书评

暂无评论

On the conceptualization of total disturbance and its profound implications

引用

science China(Information sciences) 2020年第2期63卷 221-223页

作者： Sen CHEN Wenyan BAI Yu HU Yi HUANG Zhiqiang GAO Key Laboratory of Systems and Control Academy of Mathematics and Systems ScienceChinese Academy of Sciences School of Mathematical Sciences University of Chinese Academy of Sciences Beijing Aerospace Automatic Control Institute Center for Advanced Control Technologies Department of Electrical Engineering and Computer ScienceCleveland State University

Dear editor,How to deal with uncertainties and/or disturbances is a central issue pushing the development of both control science and control technology. Among various approaches, the active disturbance rejection control (ADRC) has been successfully implemented in various industrial practices because of its uniqueness in concepts, simplicity

关键词： ESO On the conceptualization of total disturbance and its profound implications

来源：评论

学校读者我要写书评

暂无评论

Fault Diagnosis Strategy for Few Shot Industrial Process Based on Data Augmentation and Depth Information Extraction

SSRN

引用

SSRN 2022年

作者： Tian, Ying Xiang, Xin Peng, Xin Zhong, Yin Zhang, Wei Department of control science and Engineering School of Optical-Electrical and Computer Engineering University of Shanghai for Science and Technology Shanghai200093 China Key Laboratory of Advanced Control and Optimization for Chemical Processes Ministry of Education East China University of Science and Technology Shanghai200237 China

Data based intelligent fault diagnosis method is an important tool for ensuring the stability of industrial process. However, in the actual industrial process, due to the difficulty of feature extraction and the lack of labeled fault data, it is hard to form a fault diagnosis model with good performance. To address this problem, the Self-Attention embedded Generative Adversarial Networks combined with the Residual Network(SAGAN-ResNet) is proposed in this study. Firstly, to handle the lack of the fault data, the data enhancement method which consists of Self-Attention embedded generator and discriminator is adopted. Then, to extract the feature for better diagnosis performance, based on the augmented training dataset, Residual Network (ResNet) is introduced. Finally, the proposed method is compared with others, the results show that proposed method has advantages in the case of complex process fault diagnosis with few shot industrial data. © 2022, The Authors. All rights reserved.

关键词： Failure analysis

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：