检索结果-内蒙古大学图书馆

IEEE International Congress on Electronics, Electrical engineering and Computing (INTERCON)

作者： Mauricio Salazar Danushka Bandara Gissella Bejarano Digital Humanities Laboratory Pontificia Universidad Católica del Perú Lima Perú Dept. of Comp. Sci. and Engineering Fairfield University Fairfield Connecticut Department of Computer Science Marist College Poughkeepsie NY USA

ISBN: (数字)9798350378344

ISBN: (纸本)9798350378351

Due to the improvement of sign language (SL) recognition models, more work has started to build upon them to create end-user platforms such as video-based dictionaries. Although most of them base their design on cloud services or APIs, this approach can become expensive, especially when working with such substantial dimensional data as videos. Other realities, such as regions with low internet connection and few economic resources, can benefit from the offline alternative for deploying these systems. However, models deployed in websites' front-end or edge devices have some constraints on their size. Therefore, our work evaluates two model compression techniques to enhance the performance of on-device inference for a SL recognition model. We test knowledge quantization and distillation on three datasets varied in size, ultimately deploying and evaluating them on a locally hosted web page. To the best of our knowledge, this is the first evaluation of a quantized and distilled transformer-based SL recognition model for on-device browser inference. Furthermore, the combination of both compression techniques achieves a size reduction of up to 17.83 times and a speedup of 5.9 times the baseline.

关键词： Performance evaluation Sign language Adaptation models Quantization (signal) Biological system modeling Computational modeling Web pages Transformers Browsers Load modeling

来源：评论

学校读者我要写书评

暂无评论

Text-to-Image Diffusion Models can be Easily Backdoored through Multimodal Data Poisoning

arXiv

引用

arXiv 2023年

作者： Zhai, Shengfang Dong, Yinpeng Shen, Qingni Pu, Shi Fang, Yuejian Su, Hang School of Software and Microelectronics Peking University China Department of Computer Science & Technology Tsinghua University China ShengShu Beijing China School of Software and Microelectronics National Engineering Research Center for Software Engineering PKU-OCTA Laboratory for Blockchain and Privacy Computing Peking University Beijing100871 China Dept. of Comp. Sci. and Tech. Institute for AI Tsinghua-Bosch Joint ML Center THBI Lab BNRist Center Tsinghua University Beijing100084 China

With the help of conditioning mechanisms, the state-of-the-art diffusion models have achieved tremendous success in guided image generation, particularly in text-to-image synthesis. To gain a better understanding of the training process and potential risks of text-to-image synthesis, we perform a systematic investigation of backdoor attack on text-to-image diffusion models and propose BadT2I, a general multimodal backdoor attack framework that tampers with image synthesis in diverse semantic levels. Specifically, we perform backdoor attacks on three levels of the vision semantics: Pixel-Backdoor, Object-Backdoor and Style-Backdoor. By utilizing a regularization loss, our methods efficiently inject backdoors into a large-scale text-to-image diffusion model while preserving its utility with benign inputs. We conduct empirical experiments on Stable Diffusion, the widely-used text-to-image diffusion model, demonstrating that the large-scale diffusion model can be easily backdoored within a few fine-tuning steps. We conduct additional experiments to explore the impact of different types of textual triggers, as well as the backdoor persistence during further training, providing insights for the development of backdoor defense methods. Besides, our investigation may contribute to the copyright protection of text-to-image models in the future. Our Code: https://***/sf-zhai/BadT2I. Copyright © 2023, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Clinical Trial Search Using Lucene and UMLS 30

Clinical Trial Search Using Lucene and UMLS

引用

30th Text REtrieval Conference, TREC 2021

作者： Ji, Yanqing Tian, Yun Ying, Hao Tran, John Department of Electrical and Computer Engineering Gonzaga University SpokaneWA United States Department of Computer Sci. and Electrical Eng. Eastern Washington University SpokaneWA United States Department of Electrical and Computer Engineering Wayne State University DetroitMI United States Department of Psychiatry University of California San Francisco at Fresno FresnoCA United States

We approached the clinical trial search task of the 2021 TREC Clinical Trials Track as a query problem. A query (also known as a topic in 2021 TREC) is the free text description of a patient record, while the corpus is a large set of clinical trials descriptions. A commercial search engine, Lucene, was utilized for this clinical trial matching process. Namely, given a query, the system searches in the corpus and returns a subset of clinical trials with specific requirements. In this study, Unified Medical Language System (UMLS) was employed to convert the free text of both topics and clinical trials to more meaningful biomedical concepts, each of which is represented as a Concept Unique Identifier (CUI). An expansion technique based on Medical Subject Headings (MeSH) was used to expand all the condition terms for each clinical trial to their child terms. © TREC *** rights reserved.

关键词： Search engines

来源：评论

学校读者我要写书评

暂无评论

Securing the Internet of Things via Blockchain-Aided Smart Contracts 13

Securing the Internet of Things via Blockchain-Aided Smart C...

引用

13th International Conference on Information and Knowledge Technology, IKT 2022

作者： Zanjani, S. Mohammadali Shahinzadeh, Hossein Moradi, Jalal Rezaei, Zohreh Kaviani-Baghbaderani, Bahareh Tanwar, Sudeep Najafabad Branch Islamic Azad University Department of Electrical Engineering Najafabad Iran Smart Microgrid Research Center Najafabad Branch Islamic Azad University Najafabad Iran Amirkabir University of Technology Department of Electrical Engineering Tehran Iran Iran Grid Secure Operation Research Institute Amirkabir University of Technology Tehran Iran School of Electrical and Computer Engineering College of Engineering University of Tehran Tehran Iran Institute of Technology Nirma University Department of Comp. Sci. & Engg Ahmedabad382481 India

ISBN: (纸本)9798350320626

The Internet of Things (IoT) paradigm is a relatively novel concept that is rapidly and steadily growing in modern wireless communication models. In IoT, many of the objects in the surrounding environment of us are networked in one or more forms. Sensor network technologies are increasing to meet this new challenge where invisible information and communication systems are embedded in our surroundings. This produces a large amount of information that must be acquired, transferred, stored, sorted, processed, and presented in an integrated, efficient, sensible, and easily interpretable format. Therefore, this interaction and cooperation between humans and smart devices create new challenges for exchanging information security. On the other hand, the lack of appropriate and strong security services, arrangements, and measures can have a negative or even detrimental effect on the objective acceptance of the IoT paradigm. This paper delves into the impacts of the message queuing telemetry transport (MQTT) protocol, which is a widely-applied protocol in the implemented IoT schemes. This protocol does not guarantee secure and reliable native authentication mechanisms requested by developers. Thus, this paper proposes an authentication scheme via blockchain-based smart contracts for MQTT. This proposal enables the authentication of devices in the IoT network by preserving client security and privacy to ensure trust and accountability through Ethereum-based smart contracts. This review article intends to express the security challenges of this issue as well as its proposed solution in the form of a general description of the work. © 2022 IEEE.

关键词： Smart contract

来源：评论

学校读者我要写书评

暂无评论

A Non Linear System Controlling using Constrained based Genetic Algorithm 5

A Non Linear System Controlling using Constrained based Gene...

引用

5th International Conference on Smart Electronics and Communication, ICOSEC 2024

作者： Vidyarthi, Vrij Mohan Mishra, Deepali Mittal, Amit Bandyopadhyay, Krishnan Pundhir, Shivali Varshney, Niharika Tula's Institute Department of Eee U.K. Dehradun India Invertis University Deptt. of Computer Application Uttar Pradesh Bareilly India Graphic Era Hill University Dept of Allied Sciences Bhimtal Campus India Dayananda Sagar College of Engineering Deptt. of Medical Electronics Engg. Bengaluru560078 India Tula's Institute Deptt. of Computer Sci. & Engg U.K. Dehradun India Graphic Era Deemed to be University Deptt. of Electrical Engg. Dehradun India

ISBN: (纸本)9798331504403

This article presents the controlling of non-linear system with artificial intelligence (AI) technique. The non- linear system is represented by second order function and genetic algorithm will be used as AI technique. The genetic algorithm involves the three steps: reproduction, crossover, and mutation. The non-linear system is diverging in nature which is showing the non-bounded system and unstable. If the constraints are included in the mathematical expression of the second order system, the performance parameters can be improved. The second order nonlinear system including the constraints is trained by using genetic algorithm (GA). After the implementation of GA, non linear system is converging fast to bounded value or finite region. © 2024 IEEE.

关键词： Consensus algorithm

来源：评论

学校读者我要写书评

暂无评论

Cycle-object consistency for image-to-image domain adaptation

引用

Pattern Recognition 2023年第1期138卷

作者： Lin, Che-Tsung Kew, Jie-Long Chan, Chee Seng Lai, Shang-Hong Zach, Christopher Department of Electrical Engineering Chalmers University of Technology Sweden CISiP Faculty of Comp Sci. and Info. Tech. Universiti Malaya Malaysia Microsoft AI R&D Center Taiwan Department of Computer Science National Tsing Hua University Taiwan

Recent advances in generative adversarial networks (GANs) have been proven effective in performing domain adaptation for object detectors through data augmentation. While GANs are exceptionally successful, those methods that can preserve objects well in the image-to-image translation task usually require an auxiliary task, such as semantic segmentation to prevent the image content from being too distorted. However, pixel-level annotations are difficult to obtain in practice. Alternatively, instance-aware image-translation model treats object instances and background separately. Yet, it requires object detectors at test time, assuming that off-the-shelf detectors work well in both domains. In this work, we present AugGAN-Det, which introduces Cycle-object Consistency (CoCo) loss to generate instance-aware translated images across complex domains. The object detector of the target domain is directly leveraged in generator training and guides the preserved objects in the translated images to carry target-domain appearances. Compared to previous models, which e.g., require pixel-level semantic segmentation to force the latent distribution to be object-preserving, this work only needs bounding box annotations which are significantly easier to acquire. Next, as to the instance-aware GAN models, our model, AugGAN-Det, internalizes global and object style-transfer without explicitly aligning the instance features. Most importantly, a detector is not required at test time. Experimental results demonstrate that our model outperforms recent object-preserving and instance-level models and achieves state-of-the-art detection accuracy and visual perceptual quality. © 2023 Elsevier Ltd

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Cancer Death Cases Forecasting using Supervised Machine Learning

Cancer Death Cases Forecasting using Supervised Machine Lear...

引用

Electronics and Sustainable Communication Systems (ICESC), 2020 International Conference on

作者： Sheshang Degadwala Dhairya Vyas Ankita Kothari Urmila Khunt Computer Engineering Department Sigma University Vadodara Gujarat The Maharaja Sayajirao University of Baroda Vadodara Gujarat India Dept. of Comp. Sci. & Eng. Krishna School Of Emerging Technology And Applied Research KPGU Vadodara Computer Science Engineering ITM vocational University Vadodara

In India, like in the rest of the world, cancer is a major killer. This research objective is to predict cancer mortality in India, using supervised machine learning methods. Cancer mortality rates in India between 1990 and 2017 are provided by age group, gender, and region using data from the Global Burden of Disease Study. We employ three distinct supervised learning algorithms—linear regression, decision tree regression, and random forest regression—after performing data preprocessing, which includes missing value imputation and feature engineering. Using a variety of criteria, we analyze the effectiveness of these models and conclude that the random forest regression model is superior to the other two. The scope of research is provide a long-term prediction of cancer mortality in India using the best model so it will help health department to work on it. Our research has implications for policymakers and healthcare providers in India, where it may inform efforts to reduce cancer rates and improve cancer care.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Deep Learning for Disease Assessment

Deep Learning for Disease Assessment

引用

Communication, Computing and Energy Efficient Technologies (I3CEET), International Conference on

作者： Anjani Kumar Rai Chandni Manral Kawalpreet Haider Alabdeli Achilov Docent Naveen Naval Department of Computer Engineering and Applications GLA University Mathura India College of Nursing Graphic Era Hill University India Department of Computer Science Engineering Chandigarh Engineering College Chandigarh Group of Colleges Mohali Punjab India Department of Computers Techniques Engineering College of Technical Engineering The Islamic University of Al Diwaniyah Al Diwaniyah Iraq Department of Physics and Chemistry Tashkent Institute of Irrigation and Agricultural Mechanization Engineers" National Research University Tashkent Uzbekistan Western Caspian University Scientific Researcher Baku Azerbaijan Department of Humanities & Social Sci. Graphic Era Deemed to be University Dehradun India

ISBN: (数字)9798331541583

ISBN: (纸本)9798331541590

To provide readers a historical overview of cancer categorization systems, we begin this article by going over the fundamentals of cancer diagnosis. This covers the procedures involved in diagnosing cancer as well as the common categorization systems used by medical professionals. Asymmetry, edge, color and diameter, Menzies, seven-point detection, and pattern analysis are some of these techniques. Physicians often use them to identify cancer, although it’s debatable if they improve performance. Furthermore, all target audiences are given attention in accordance with the fundamental assessment criteria. The following criteria are used: F1 score, precision, specificity, sensitivity, accuracy, Dice coefficient, accuracy, and mean accuracy. It is thought that because earlier techniques were ineffective, more sophisticated cancer detection techniques need to be used. He makes his Python scripts available to interested readers for each strategy so they may try out the methods on their own diagnostic issues. The book’s last section provides an overview of deep learning models that have been successfully applied to various cancer therapy modalities. We shall restrict our discussion to breast, lung, brain, and skin cancers due to the size of this book. This bibliographic study aims to provide sci.ntists a thorough understanding of the most recent advancements in neural network and deep learning methods for cancer detections.

关键词： Deep learning Accuracy Lungs Cancer detection Skin Telecommunication computing Convolutional neural networks Standards Cancer Skin cancer

来源：评论

学校读者我要写书评

暂无评论

Econometric and AI-Based Modelling of Nigeria's Interest Rates Based on Fisher Equation

Econometric and AI-Based Modelling of Nigeria's Interest Rat...

引用

International Symposium on Digital Forensic and Security (ISDFS)

作者： Sagiru Mati Salim Jibrin Danbatta Asaf Varol Ahad Nasab Abdullahi Garba Usman Berna Uzun Ahmad Muhammad Operational Research Centre in Healthcare Near East University TRNC Mersin 10 Nicosia Turkey Department of Economics Yusuf Maitama University Kano Nigeria Software Engineering Department Uskudar University Istanbul Türkiye Dept. of Engineering Mgmt. & Tech. College of Engineering and Comp. Sci. The University of Tennessee at Chattanooga Chattanooga TN United States Department of Physics and Materials Sciences College of Arts and Sciences Qatar University Doha Qatar

ISBN: (数字)9798350330366

ISBN: (纸本)9798350330373

In this study, linear and non-linear models are employed to model the Central Bank Rate (CBR), Deposit Interest Rate (DIR), Lending Rate (LER), and Treasury Bill Rate (TBR). The choice of variables is guided by the Fisher Equation (FE). The linear models range from Autoregressive Integrated Moving Average (ARIMA), Simple Linear Regression (SLR), Autoregressive Distributed Lag (ARDL) model, Vector Autoregressive (VAR) model, Bayesian VAR (BVAR) to Vector Error Correction Model (VECM). The nonlinear models include Artificial Neural Network (ANN) and Adaptive Neuro-Fuzzy Inference System (ANFIS). The results show that VECM outperforms other models in the testing sub-sample, while ANFIS produces the highest forecast accuracy in the training sub-sample. Ensembling the VECM and ANFIS produces the best forecast accuracy for both the sub-samples. The results of this study can serve as a reference for modeling interest rates and provide the benchmark for investment decisions and macroeconomic policy designs.

关键词： Training Adaptation models Reactive power Economic indicators Biological system modeling Predictive models Prediction algorithms

来源：评论

学校读者我要写书评

暂无评论

CMOS-based Single-Cycle In-Memory XOR/XNOR

arXiv

引用

arXiv 2023年

作者： Alam, Shamiul Hutchins, Jack Shukla, Nikhil Asifuzzaman, Kazi Aziz, Ahmedullah Dept. of Electrical Eng. and Computer Sci. University of Tennessee KnoxvilleTN37996 United States Department of Electrical and Computer Engineering University of Virginia CharlottesvilleVA United States Oak Ridge National Laboratory Oak RidgeTN United States

Big data applications are on the rise, and so is the number of data centers. The ever-increasing massive data pool needs to be periodically backed up in a secure environment. Moreover, a massive amount of securely backed-up data is required for training binary convolutional neural networks for image classification. XOR and XNOR operations are essential for large-scale data copy verification, encryption, and classification algorithms. The disproportionate speed of existing compute and memory units makes the von Neumann architecture inefficient to perform these Boolean operations. Compute-in-memory (CiM) has proved to be an optimum approach for such bulk computations. The existing CiM-based XOR/XNOR techniques either require multiple cycles for computing or add to the complexity of the fabrication process. Here, we propose a CMOS-based hardware topology for single-cycle in-memory XOR/XNOR operations. Our design provides at least 2× improvement in the latency compared with other existing CMOS-compatible solutions. We verify the proposed system through circuit/system-level simulations and evaluate its robustness using a 5000-point Monte Carlo variation analysis. This all-CMOS design paves the way for practical implementation of CiM XOR/XNOR at scaled technology nodes. © 2023, CC BY.

关键词： Cryptography

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：