检索结果-内蒙古大学图书馆

Annual Conference on Privacy, Security and Trust, PST

作者： Weimin Zhao Qusay H. Mahmoud Sanaa Alwidian Dept. of Electrical Computer and Software Engineering Ontario Tech University Oshawa ON Canada

Deep learning models are being utilized and further developed in many application domains, but challenges still exist regarding their interpretability and consistency. Interpretability is important to provide users with transparent information that enhances the trust between the user and the learning model. It also gives developers feedback to improve the consistency of their deep learning models. In this paper, we present a novel architectural design to embed interpretation into the architecture of the deep learning model. We apply dynamic pixel-wised weights to input images and produce a highly correlated feature map for classification. This feature map is useful for providing interpretation and transparent information about the decision-making of the deep learning model while keeping full context about the relevant feature information compared to previous interpretation algorithms. The proposed model achieved 92% accuracy for CIFAR 10 classifications without finetuning the hyperparameters. Furthermore, it achieved a 20% accuracy under 8/255 PGD adversarial attack for 100 iterations without any defense method, indicating extra natural robustness compared to other Convolutional Neural Network (CNN) models. The results demonstrate the feasibility of the proposed architecture.

关键词：

来源：评论

学校读者我要写书评

暂无评论

XGSwap: eXtreme Gradient boosting Swap for Routing in NISQ Devices

arXiv

引用

arXiv 2024年

作者： Waring, Jean-Baptiste Pere, Christophe Le Beux, Sébastien Dept. of Electrical & Computer Engineering Concordia University MontréalQC Canada Dept. of Computer Science & Software Engineering INTRIQ Laval University QuébecQC Canada PINQ2

In the current landscape of noisy intermediate-scale quantum (NISQ) computing, the inherent noise presents significant challenges to achieving high-fidelity long-range entanglement. Furthermore, this challenge is amplified by the limited connectivity of current superconducting devices, necessitating state permutations to establish long-distance entanglement. Traditionally, graph methods are used to satisfy the coupling constraints of a given architecture by routing states along the shortest undirected path between qubits. In this work, we introduce a gradient boosting machine learning model to predict the fidelity of alternative–potentially longer–routing paths to improve fidelity. This model was trained on 4050 random CNOT gates ranging in length from 2 to 100+ qubits. The experiments were all executed on ibm_quebec, a 127-qubit IBM Quantum System One. Through more than 200+ tests run on actual hardware, our model successfully identified higher fidelity paths in approximately 24% of *** Codes 81P68 © 2024, CC BY-NC-SA.

关键词： Quantum entanglement

来源：评论

学校读者我要写书评

暂无评论

Natural Language Processing for Conversational AI: Chatbots and Virtual Assistants

Natural Language Processing for Conversational AI: Chatbots ...

引用

Interdisciplinary Approaches in Technology and Management for Social Innovation (IATMSI), IEEE Conference on

作者： Neeraj Shrivastava Pushpa Tewari S. Sujatha Srinivasa Rao Bogireddy Neeraj Varshney Vinod Sharma Dept. of Engineering Medicaps University Indore India Dept. of Vocational Studies Guru Nanak College Dhanbad India Dept. of Civil Engineering K. Ramakrishnan College of Technology Trichy India Dept. of Software Engineer Horizon Systems Inc Phoenix Arizona Dept. of CSE GLA University Mathura India Dept. of Computer Science Jiwaji University Gwalior India

ISBN: (数字)9798331521691

ISBN: (纸本)9798331521707

The frontline in developing the capabilities of chatbots and virtual assistants involves the use of NLP, enabling these machines to evolve into truly advanced, interactive machines that are able to comprehend and respond to human language with an almost unimaginable degree of accuracy. The paper will analyze some of the major NLP techniques and methodologies used for such systems: machine learning algorithms, neural networks, and semantic analysis. This would also mean that, with the help of big data and continuous learning, the work of chatbots and virtual assistants will range from answering queries to personalized recommendations. Language understanding, generation, and context management are discussed in detail, underlining how NLP improves user experience by making interactions more natural and intuitive. The challenges of ambiguous language processing, conversational context, and ensuring data privacy and security are discussed as well. Case studies and applications in real life show the impact of NLP on different sectors: customer service, healthcare, and e-commerce. The findings bring out the potential of NLP to change human-computer interaction and allow for more responsive and intelligent digital assistants.

关键词： Ethics Technological innovation Sentiment analysis Virtual assistants Semantics Speech recognition Chatbots Natural language processing User experience Text processing

来源：评论

学校读者我要写书评

暂无评论

Joint Downlink-Uplink Beamforming for Wireless Multi-Antenna Federated Learning

arXiv

引用

arXiv 2023年

作者： Zhang, Chong Dong, Min Liang, Ben Afana, Ali Ahmed, Yahia Dept. of Electrical and Computer Engineering University of Toronto Canada Ericsson Canada Canada Dept. of Electrical Computer and Software Engineering Ontario Tech University Canada

We study joint downlink-uplink beamforming design for wireless federated learning (FL) with a multi-antenna base station. Considering analog transmission over noisy channels and uplink over-the-air aggregation, we derive the global model update expression over communication rounds. We then obtain an upper bound on the expected global loss function, capturing the downlink and uplink beamforming and receiver noise effect. We propose a low-complexity joint beamforming algorithm to minimize this upper bound, which employs alternating optimization to breakdown the problem into three subproblems, each solved via closed-form gradient updates. Simulation under practical wireless system setup shows that our proposed joint beamforming design solution substantially outperforms the conventional separate-link design approach and nearly attains the performance of ideal FL with error-free communication links. Copyright © 2023, The Authors. All rights reserved.

关键词： Beamforming

来源：评论

学校读者我要写书评

暂无评论

Robust and Reliable Liveness Detection Models for Facial Recognition systems 20

Robust and Reliable Liveness Detection Models for Facial Rec...

引用

20th International Conference on Frontiers of Information Technology, FIT 2023

作者： Anjum, Haris Arshad, Usama Ali, Raja Hashim Ul Abideen, Zain Shah, Muhammad Huzaifa Khan, Talha Ali Ijaz, Ali Zeeshan Siddique, Abu Bakar Imad, Muhammad Gik Institute of Engg. Sciences & Technology Faculty of Computer Science & Engineering Topi Pakistan University of Europe for Applied Sciences Dept. of Technology & Software Engineering Berlin Germany BCMaterials Basque Center for Materials Applications & Nanostructure Spain

ISBN: (纸本)9798350395785

Facial Liveness Detection is instrumental in combating fraudulent practices and identity theft by differentiating genuine faces from forgeries. Given that facial recognition is now an integral part of many sectors like banking and law enforcement, liveness detection has become a vital aspect for maintaining the trustworthiness of these applications. Unfortunately, the utility of prevalent facial liveness detection models is constrained by their lack of generalizability to novel, unseen data, which presents a substantial impediment to practical applications. To mitigate this issue, our research endeavored to construct a reliable and precise facial liveness detection model that can effectively navigate this challenge and ensure consistent performance across diverse datasets. The devised model exhibited excellent accuracy, which attests to its robustness. This research underscores the advantages of employing deep learning methods such as Convolutional Neural Networks (CNNs) as they significantly enhance the model's accuracy. We intend for our findings to drive further research and progress in this critical field, by laying the groundwork for future work that integrates multiple liveness detection methods and collects more varied training data to develop an even more resilient and efficient system. © 2023 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Multi-Point Positional Insertion Tuning for Small Object Detection

Multi-Point Positional Insertion Tuning for Small Object Det...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Kanoko Goto Takumi Karasawa Takumi Hirose Rei Kawakami Nakamasa Inoue Dept. of Computer Science Institute of Science Tokyo Tokyo Japan Dept. of Systems and Control Engineering Institute of Science Tokyo Tokyo Japan

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Small object detection aims to localize and classify small objects within images. With recent advances in large-scale vision-language pretraining, finetuning pretrained object detection models has emerged as a promising approach. However, finetuning large models is computationally and memory expensive. To address this issue, this paper introduces multi-point positional insertion (MPI) tuning, a parameter-efficient finetuning (PEFT) method for small object detection. Specifically, MPI incorporates multiple positional embeddings into a frozen pretrained model, enabling the efficient detection of small objects by providing precise positional information to latent features. Through experiments, we demonstrated the effectiveness of the proposed method on the SODA-D dataset. MPI performed comparably to conventional PEFT methods, including CoOp and VPT, while significantly reducing the number of parameters that need to be tuned.

关键词： Computational modeling Object detection Signal processing Feature extraction Encoding Acoustics Speech processing Tuning

来源：评论

学校读者我要写书评

暂无评论

Unveiling The Arsenal of User Data Protection Tools and Practices

Unveiling The Arsenal of User Data Protection Tools and Prac...

引用

Humanitarian Technology Conference (KHI-HTC), IEEE Karachi Section

作者： Syed Moiz Abdul Ahad Abro Mansoor Ebrahim Abdul Basit Abro Dept. of Computer Science Iqra University Main Campus Karachi Pakistan Dept. of Software Engineering Iqra University Main Campus Karachi Pakistan Dept. of Computer Science Indus University Karachi Pakistan

In our rapidly evolving digital landscape, the imperative of safeguarding personal data has surged in significance. As lives increasingly intertwine with digital technologies, personal information has grown markedly, parallel to the escalating threats posed by cyberattacks and data breaches. This research paper provides a comprehensive exploration of data security and privacy, aiming to impart a profound understanding of the intrinsic value of data. It meticulously examines essential methodologies such as encryption, access control, multifactor authentication, and systems, all of which collectively form the bedrock of a resilient security framework. Additionally, this study delves into the intricate regulatory landscape that governs data protection, placing particular emphasis on well-established frameworks such as the General Data Protection Regulation and the Health Insurance Portability and Accountability Act. It highlights emergent trends in data security, including the integration of artificial intelligence and machine learning for threat detection, while also addressing the distinctive challenges posed by the proliferation of Internet of Things (IoT) devices. In conclusion, this paper underscores the paramount importance of personal data protection in our digital age, offering invaluable insights and guidance for organizations and individuals endeavoring to bolster their defenses and uphold data integrity within our interconnected world.

关键词：

来源：评论

学校读者我要写书评

暂无评论

EMDA-Net: Earth Mover’s Distance (EMD) influenced Attention-aided Neural Network for Medical Image Classification

EMDA-Net: Earth Mover’s Distance (EMD) influenced Attention...

引用

IEEE International Conference on Cognitive Machine Intelligence (CogMI)

作者： Surya Majumder Kushaj Mallick Wrick Pal Somenath Chakraborty Ram Sarkar Dept. of Computer Science & Engineering Heritage Institute of Technology Kolkata India Dept. of Electrical Engineering Jadavpur University Kolkata India Dept. of Computer Science & Information Systems West Virginia University Institute of Technology West Virginia USA Dept. of Computer Science & Engineering Jadavpur University Kolkata India

ISBN: (数字)9798350386721

ISBN: (纸本)9798350386738

Medical image analysis methods need to be more generalized in order to meet the substantial problems associated with cancer detection and treatment, including skin, cervical, and breast cancer. In this work, we propose a novel lightweight convolutional neural network (CNN) architecture, named EMDA-Net (Earth Mover’s Distance (EMD) influenced Attention-aided Neural Network), specifically designed to meet the requirements of cancer subtype classification. It is built on the MobileNetV2 backbone and is equipped with hybrid statistical mechanism-aided attention mechanisms. Our model ensures high-performance while maintaining a lightweight design by leveraging the capabilities of the MobileNetV2 architecture and the attention mechanism. To improve similarity detection across data channels, we present a novel EMD influenced attention technique. We have examined datasets related to skin, cervical, and breast cancer for experimentation. Our model especially outperforms the state-of-the-art techniques, demonstrating its effectiveness in maximizing diagnosis precision for various cancer types across the datasets. The code of the proposed model will be made public at: https://***/q5DlN

关键词： Earth Attention mechanisms Codes Neural networks Skin Breast cancer Convolutional neural networks Medical diagnostic imaging Machine intelligence Image classification

来源：评论

学校读者我要写书评

暂无评论

Multi-Model Wireless Federated Learning with Downlink Beamforming

arXiv

引用

arXiv 2023年

This paper studies the design of wireless federated learning (FL) for simultaneously training multiple machine learning models. We consider round robin device-model assignment and downlink beamforming for concurrent multiple model updates. After formulating the joint downlink-uplink transmission process, we derive the per-model global update expression over communication rounds, capturing the effect of beamforming and noisy reception. To maximize the multi-model training convergence rate, we derive an upper bound on the optimality gap of the global model update and use it to formulate a multi-group multicast beamforming problem. We show that this problem can be converted to minimizing the sum of inverse received signal-to-interference-plus-noise ratios, which can be solved efficiently by projected gradient descent. Simulation shows that our proposed multi-model FL solution outperforms other alternatives, including conventional single-model sequential training and multi-model zero-forcing beamforming. Copyright © 2023, The Authors. All rights reserved.

关键词： Beamforming

来源：评论

学校读者我要写书评

暂无评论

Joint Downlink-Uplink Beamforming for Wireless Multi-Antenna Federated Learning

Joint Downlink-Uplink Beamforming for Wireless Multi-Antenna...

引用

International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt)

作者： Chong Zhang Min Dong Ben Liang Ali Afana Yahia Ahmed Dept. of Electrical and Computer Engineering University of Toronto Canada Dept. of Electrical Computer and Software Engineering Ontario Tech University Canada Ericsson Canada Canada

关键词：

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：