检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Liu, Qinghui Tang, Zhiyi Sch. Comp. Sci. Beijing Inst. Tech. Beijing100081 China

We study a class of Schrödinger operators Hm,λwith generalized Thue-Morse potential that generated by the substitution τ(a) = ambm, τ(b) = bmamon two symbol alphabet Σ = {a, b} for integer m ≥ 2 and coupling λ > 0. We show that (Formula presented) where σ(Hm,λ) is the spectrum of Hm,λ, Λ2= 2, and for m > 2, Λm= m, if m ≡ 0 mod 4;Λm= m - 3, if m ≡ 1 mod 4;Λm= m - 2, if m ≡ 2 mod 4;Λm= m - 1, if m ≡ 3 mod 4. This implies that dimHσ(Hm,λ) tends to 1 as m tends to *** Codes 28A78, 81Q10, 47B80 © 2022, CC BY.

关键词： Morse potential

来源：评论

学校读者我要写书评

暂无评论

A Kubernetes-Powered Personalized Federated Learning Platform for Resource-Constrained Internet of Medical Things

A Kubernetes-Powered Personalized Federated Learning Platfor...

引用

International Wireless Communications and Mobile comp.ting Conference, IWCMC

作者： Meng-Hsuan Lin Shun-Ren Yang Hsin-Yin Chiang Chu-Chun Chang Phone Lin Dept. of Comp. Sci Inst. of Info. Sys. & App. and Inst. of Commun. Eng. National Tsing Hua University Taiwan Cephalgo SAS 8 rue des Veaux Strasbourg France Dept. of Comp. Sci. & Info. Eng Grad. Inst. of Biomed. Elec. & Bioinform. Grad. Inst. of Net. & MM Inst. of Med. Device & Imaging and IoT Research Center National Taiwan University Taiwan

ISBN: (数字)9798350361261

ISBN: (纸本)9798350361278

Federated Learning (FL) is a distributed machine learning scheme that trains a global model across multiple end devices while protecting user privacy by keeping data locally, which has been shown to be useful for smart healthcare. By integrating virtualized container technology into edge platforms, flexible comp.ting resources can be made available to Internet of Medical Things (IoMT) devices, enabling their participation in FL. Some studies have utilized the advantages of Kubernetes (K8s) for fast container deployment, offering general FL platforms for developers to rapidly deploy models. However, there are currently no FL development platforms suitable for smart healthcare, which exhibits unique challenges of personalized medical data and large-sized models for resource-limited IoMT devices. In this paper, we propose a Kubernetes-powered personalized FL (PFL) platform for resource-constrained IoMT. This platform, which incorporates personalized strategies to capture individual features and model comp.ession mechanisms to reduce model size, allows model developers to formulate PFL training tasks for healthcare users to acquire personalized and comp.essed models, deployable on their devices. Via establishing a real testbed and deploying a demonstrating training task, our experiments verify the platform’s ability to support the development of personalized and comp.essed models.

关键词： Training Wireless communication Federated learning comp.tational modeling Distributed databases Internet of Medical Things Containers

来源：评论

学校读者我要写书评

暂无评论

MouseScholar: Evaluating an Image+Text Search System for Biocuration

MouseScholar: Evaluating an Image+Text Search System for Bio...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Juan Trelles Trabucco Carla Floricel Cecilia Arighi Hagit Shatkay Daniela Raciti Martin Ringwald G. Elisabeta Marai Dept. Comp. Sci. University of Illinois Chicago Chicago IL Dept. Comp. Info. Sci. University of Delaware Newark DE Div. Biology Biological Eng. California Inst. Technology Pasadena California The Jackson Laboratory Bar Harbor Maine

Biocuration is the process of analyzing biological or biomedical articles to organize biological data into data repositories using taxonomies and ontologies. Due to the expanding number of articles and the relatively small number of biocurators, automation is desired to improve the workflow of assessing articles worth curating. As figures convey essential information, automatically integrating images may improve curation. In this work, we inst.ntiate and evaluate a first-in-kind, hybrid image+text document search system for biocuration. The system, MouseScholar, leverages an image modality taxonomy derived in collaboration with biocurators, in addition to figure segmentation, and classifiers comp.nents as a back-end and a streamlined front-end interface to search and present document results. We formally evaluated the system with ten biocurators on a mouse genome informatics biocuration dataset and collected feedback. The results demonstrate the benefits of blending text and image information when presenting sci.ntific articles for biocuration.

关键词：

来源：评论

学校读者我要写书评

暂无评论

On Topology Optimization Enhancements for Thin-Film Lithium Niobate

On Topology Optimization Enhancements for Thin-Film Lithium ...

引用

LEOS Summer Topical Meetings

作者： Michael J. Probst Jacob M. Hiesener Leticia Magalhaes C. J. Xin Benjamin Szamosfalvi Ryan M. Camacho Marko Lončar Stephen E. Ralph Sch. of Elec. and Comp. Eng. Georgia Inst. of Technology Atlanta USA Sch. of Eng. and Apl. Sci. Harvard University Cambridge USA Sch. of Elec. and Comp. Eng. Brigham Young University Provo USA

We present topology optimization enhancements for thin-film lithium niobate, enabling optimization for structures with nonvertical sidewalls and anisotropic media. We validate our enhancements by optimizing and measur... 详细信息

ISBN: (数字)9798350393873

ISBN: (纸本)9798350393880

关键词： Demultiplexing Three-dimensional displays Lithium niobate Couplers Photonic crystals Media Topology

来源：评论

学校读者我要写书评

暂无评论

Information Hiding in Images Using Neural Network 8

Information Hiding in Images Using Neural Network

引用

8th International Conference on comp.ter sci.nce and Engineering, UBMK 2023

作者： Daiyrbayeva, Elmira Yerimbetova, Aigerim Maratov, Zhanbolat Sakenov, Bakzhan Institute of Information and Computing Technologies of the Science Committee The Ministry of Science and Higher Education of the Republic of Kazakhstan Satbayev University Academy of Logistics and Transport Almaty Kazakhstan Institute of Information and Computing Technologies of the Science Committee The Ministry of Science and Higher Education of the Republic of Kazakhstan Satbayev University Almaty Kazakhstan Inst. of Info. and Comp. Technol. of the Sci. Comm. of the Min. of Sci. and Higher Educ. of the Rep. of Kazakhstan Almaty Kazakhstan

ISBN: (纸本)9798350340815

This research work is very relevant in the field of image processing using artificial intelligence techniques. In today's world where information security plays an important role, developing methods to detect and analyze hidden data in images is a necessity. The aim of the research is to investigate and develop image processing methods using artificial intelligence techniques to improve the accuracy and efficiency of image processing. This may include creating new neural network models, exploring different data preprocessing techniques, optimizing comp.tational processes and exploring the use of generative models for image synthesis and enhancement. During the experiment, it was proved that, as a result of training the neural network, the values of the loss function decreases during the training process, which in turn means improvements in the model. The model was trained in the Kaggle service. © 2023 IEEE.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

VADtalk: An Internet of Vehicles Platform Facilitating Anomaly Detection Modeling and Deployment for Self-Driving Vehicles

VADtalk: An Internet of Vehicles Platform Facilitating Anoma...

引用

International Wireless Communications and Mobile comp.ting Conference, IWCMC

作者： Yi-Cheng Lu Shun-Ren Yang Phone Lin Chih-Wei Huang Dept. of Comp. Sci. & Inst. of Info. Sec. National Tsing Hua University Hsinchu Taiwan Dept. of Comp. Sci. & Info. Eng. National Taiwan University Taipei Taiwan Dept. of Commun. Eng. National Central University Taoyuan Taiwan

In recent years, self-driving vehicles have gradually been appearing on the road, but society has also begun to worry about the possibility of accidents caused by the anomaly self-driving system. Many researchers have begun to study the anomaly detection of self-driving vehicles, and each has proposed different detection algorithms. However, since self-driving vehicles are not yet popular, how to collect data, simulate attacks, and verify and comp.re multiple algorithms is a major obstacle to research. In this regard, we built an Internet of Vehicles platform, VADtalk, that facilitate anomaly detection modeling and deployment for self-driving vehicles. VADtalk contains programs such as anomaly detection model training and vehicle connection. When developers comp.ete model uploading and setting through the GUI, the platform will automatically collect self-driving data, train the model, and even verify the operation of the model using a self-driving simulator, and then provide the results to the developer. After the developer determines the model, VADtalk can connect the trained model with the self-driving vehicle to actually perform real-time anomaly detection on it.

关键词：

来源：评论

学校读者我要写书评

暂无评论

ANIMO: Annotation of Biomed Image Modalities

ANIMO: Annotation of Biomed Image Modalities

引用

2021 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2021

作者： Trabucco, Juan Trelles Li, Pengyuan Arighi, Cecilia Raciti, Daniela Shatkay, Hagit Marai, G. Elisabeta University of Illinois at Chicago Dept. Comp. Sci. ChicagoIL60607 United States University of Delaware Dept. Comp. Info. Sci. NewarkDE19716 United States California Inst. Technology Div. Biology Biological Eng. PasadenaCA91125 United States

ISBN: (纸本)9781665401265

Figures within biomedical articles present essential evidence of the relevance of a publication in a curation workflow. In particular, visual cues of the image modality or experimental methods can help expert curators identify relevant papers from an increasing number of publications. Automating the identification of these content-bearing images can thus be helpful in comp.ter-assisted curation. However, the paucity of labeled datasets and the specialized training required to label such images hinder the development of such tools. To address this problem, we present the design of ANIMO, a labeling system that integrates extraction and segmentation tools to ease the annotation burden. We first introduce two taxonomies of image modalities and experimental methods, derived in collaboration with curators. On the back-end of the system, we process batches of documents and create a labeling task per document. At the front-end, expert curators can access these tasks through a web interface and access the article of interest. We describe the evaluation of this system by a group of biocurators, and the human factor lessons learned from this interdisci.linary experience. © 2021 IEEE.

关键词： Image annotation

来源：评论

学校读者我要写书评

暂无评论

A 15-nW 14-ppm/°C 1.18 V startup-less bandgap-based voltage regulator

A 15-nW 14-ppm/°C 1.18 V startup-less bandgap-based voltage...

引用

IEEE International Symposium on Circuits and Systems (ISCAS)

作者： Kapil Jainwal Kalyan Kota Kumar Gajendranath Chowdary Shouri Chatterjee Electrical Eng. and Comp. Sci. Indian Inst. of Tech. Bhilai Raipur India Electrical Eng. Indian Inst. of Tech. Delhi New Delhi India Electrical Eng. Indian Inst. of Tech. Hyderabad Telangana India

We present an ultra-low-power band-gap reference-based regulator and a unique method of generating a proportional-to-absolute-temperature (PTAT) voltage and a PTAT current. The PTAT current is used to generate a BJT comp.ementary-to-absolute-temperature (CTAT) voltage. This CTAT voltage and PTAT voltage are added to generate a regulated output voltage in the negative feedback loop to drive the load. The output of the proposed regulator is temperature stable from −20° C to 85° C with a mean variation of less than 20 ppm/° C. The proposed technique eliminates the need for an additional start-up circuitry, which is mandatory for conventional architectures. The 1.18 V output voltage regulator, designed in a 180-nm CMOS process, has a simulated line regulation of 0.1 %/V with a power consumption of 15 nW at room temperature. The proposed regulator can drive the load from 0 to 30 mA.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Jailbreaking LLMs’ Safeguard with Universal Magic Words for Text Embedding Models

arXiv

引用

arXiv 2025年

作者： Liang, Haoyu Sun, Youran Cai, Yunfeng Zhu, Jun Zhang, Bo Dept. of Comp. Sci. and Tech. Inst. for AI Tsinghua-Bosch Joint ML Center THBI Lab BNRist Center Tsinghua University Beijing China Dept. of Math. Sci Tsinghua University Beijing China Microsoft Research Asia China Beijing China

The security issue of large language models (LLMs) has gained wide attention recently, with various defense mechanisms developed to prevent harmful output, among which safeguards based on text embedding models serve as a fundamental defense. Through testing, we discover that the output distribution of text embedding models is severely biased with a large mean. Inspired by this observation, we propose novel efficient methods to search for universal magic words that attack text embedding models. Universal magic words as suffixes can shift the embedding of any text towards the bias direction, thus manipulating the similarity of any text pair and misleading safeguards. Attackers can jailbreak the safeguards by appending magic words to user prompts and requiring LLMs to end answers with magic words. To eradicate this security risk, we also propose defense methods against.such attacks, which can correct the bias of text embeddings and improve downstream performance in a train-free manner. © 2025, CC BY.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

A note on circular wirelength for hypercubes

arXiv

引用

arXiv 2021年

作者： Liu, Qinghui Tang, Zhiyi Sch. Comp. Sci. Beijing Inst. Tech. Beijing100081 China

MSC Codes 05C60, 68R10We study embeddings of the n-dimensional hypercube into the circuit with 2n vertices. We prove that the circular wirelength attains minimum by gray coding, which is called the CT conjecture by Chavez and Trapp (Discrete Applied Mathematics, 1998). This problem had claimed to be settled by Ching-Jung Guu in her doctor dissertation "The circular wirelength problem for hypercubes" (University of California, Riverside, 1997). Many people argue there are gaps in her proof. We eliminate gaps in her dissertation. © 2021, CC BY.

关键词： Geometry

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：