检索结果-内蒙古大学图书馆

High-Temperature Strong Nonreciprocal Thermal Radiation from Semiconductors

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Nabavi, Bardia Ghalekohneh, Sina Jafari Shayegan, Komron J. Tervo, Eric J. Atwater, Harry Zhao, Bo Department of Mechanical and Aerospace Engineering University of Houston HoustonTX77204 United States Thomas J. Watson Laboratory of Applied Physics California Institute of Technology PasadenaCA91125 United States Department of Electrical Engineering California Institute of Technology PasadenaCA91125 United States Department of Electrical and Computer Engineering University of Wisconsin-Madison MadisonWI53706 United States Department of Mechanical Engineering University of Wisconsin-Madison MadisonWI53706 United States

Nonreciprocal thermal emitters that break the conventional Kirchhoff's law allow independent control of emissivity and absorptivity and promise exciting new functionalities in controlling heat flow for thermal and energy applications. In enabling some of these applications, nonreciprocal thermal emitters will unavoidably need to serve as hot emitters. Leveraging magneto-optical effects, degenerate semiconductors have been demonstrated as a promising optical material platform for nonreciprocal thermal radiation. However, existing modeling and experimental efforts are limited to near room temperature ( © 2025, CC BY.

关键词： Temperature distribution

Pioneering Privacy-First Approaches in IoT Healthcare with Secure Wearable Monitoring Systems

学校读者我要写书评

暂无评论

Pioneering Privacy-First Approaches in IoT Healthcare with S...

Intelligent Control, Computing and Communications (IC3), International Conference on

作者： Rusul Najm Neeta Awasthy E.N.Vijaya Kumari Sorabh Lakhannal Tarunika Sharma Anurag Singh Bh. Prashanthi Radiology Techniques Department College of Medical Technology The Islamic University Najaf Iraq Department of Computer Science and Engineering GL Bajaj Group of Institutions Mathura U.P. India Department of Computer Science and Engineering MLR Institute of Technology Hyderabad Telangana India Lovely Professional University Phagwara India Department of Applied Sciences New Horizon College of Engineering Bangalore India Department of Electrical Electronics & Communication Engineering Galgotias university Gerater Noida Uttar Pradesh India Department of CSE GRIET Hyderabad Telangana India

ISBN: (数字)9798331527495

ISBN: (纸本)9798331527501

In the world of IoT healthcare right now, protecting the privacy of patients' medical data collected by wearable tracking devices is an urgent and important issue. It is very important that this condition be met. The Secure Health Data Aggregation and Homomorphic Encryption (SHAHE) method is being thought about as a possible alternative while this investigation is going on. We thought about the need for processing, data collection, and access control as we worked on the SHAHE approach. We can fully reach this goal by using homomorphic cryptography and safe aggregation methods. The main goal of the program is to protect people's privacy by making data value analysis better and reducing unauthorized access to important health information. This is the reason why the program was made.

关键词： Access control Data privacy Privacy Process control Medical services Data aggregation Biomedical monitoring Homomorphic encryption Wearable devices Monitoring

Predicting the Parabolic Growth Rate Constant for High-Temperature Oxidation of Steels Using Machine Learning Models

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Aghaeian, Soroush Nourouzi, Farshid Sloof, Wim G. Mol, Arjan J.M.C. Böttger, Amarante J. Department of Materials Science and Engineering Faculty of Mechanical Maritime and Materials Delft University of Technology Mekelweg 2 Delft2628 CD Netherlands Department of Electrical Sustainable Energy Faculty of Electrical Engineering Mathematics and Computer Science Delft University of Technology Mekelweg 4 Delft2628 CD Netherlands

Predicting the high-temperature oxidation kinetics of different steel grades is a complex challenge that has not been fully addressed yet. In this work, a data analytics approach is presented to predict the parabolic growth rate constant (kp) of high-temperature oxidation of steels using published data. Experimental data are collected for more than 70 different steel grades exposed to temperatures between 500 and 1200 C in air. © 2023, The Authors. All rights reserved.

关键词： Forecasting

A new method for determining Wasserstein 1 optimal transport maps from Kantorovich potentials, with deep learning applications

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Milne, Tristan Bilocq, Étienne Nachman, Adrian University of Toronto Department of Mathematics Canada University of Toronto The Edward S. Rogers Sr. Department of Electrical and Computer Engineering Canada

Wasserstein 1 optimal transport maps provide a natural correspondence between points from two probability distributions, µ and ν, which is useful in many applications. Available algorithms for computing these maps do not appear to scale well to high dimensions. In deep learning applications, efficient algorithms have been developed for approximating solutions of the dual problem, known as Kantorovich potentials, using neural networks (e.g. [Gulrajani et al., 2017]). Importantly, such algorithms work well in high dimensions. In this paper we present an approach towards computing Wasserstein 1 optimal transport maps that relies only on Kantorovich potentials. In general, a Wasserstein 1 optimal transport map is not unique and is not computable from a potential alone. Our main result is to prove that if µ has a density and ν is supported on a submanifold of codimension at least 2, an optimal transport map is unique and can be written explicitly in terms of a potential. These assumptions are natural in many image processing contexts and other applications. When the Kantorovich potential is only known approximately, our result motivates an iterative procedure wherein data is moved in optimal directions and with the correct average displacement. Since this provides an approach for transforming one distribution to another, it can be used as a multipurpose algorithm for various transport problems;we demonstrate through several proof of concept experiments that this algorithm successfully performs various imaging tasks, such as denoising, generation, translation and deblurring, which normally require specialized *** Codes 49Q22 © 2022, CC BY.

关键词： Iterative methods

Sharpness-Aware Training for Free

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Du, Jiawei Zhou, Daquan Feng, Jiashi Tan, Vincent Y.F. Zhou, Joey Tianyi A*STAR Singapore Department of Electrical and Computer Engineering National University of Singapore Singapore ByteDance China Department of Mathematics National University of Singapore Singapore

Modern deep neural networks (DNNs) have achieved state-of-the-art performances but are typically over-parameterized. The over-parameterization may result in undesirably large generalization error in the absence of other customized training strategies. Recently, a line of research under the name of Sharpness-Aware Minimization (SAM) has shown that minimizing a sharpness measure, which reflects the geometry of the loss landscape, can significantly reduce the generalization error. However, SAM-like methods incur a two-fold computational overhead of the given base optimizer (e.g. SGD) for approximating the sharpness measure. In this paper, we propose Sharpness-Aware Training for Free, or SAF, which mitigates the sharp landscape at almost zero additional computational cost over the base optimizer. Intuitively, SAF achieves this by avoiding sudden drops in the loss in the sharp local minima throughout the trajectory of the updates of the weights. Specifically, we suggest a novel trajectory loss, based on the KL-divergence between the outputs of DNNs with the current weights and past weights, as a replacement of the SAM's sharpness measure. This loss captures the rate of change of the training loss along the model's update trajectory. By minimizing it, SAF ensures the convergence to a flat minimum with improved generalization capabilities. Extensive empirical results show that SAF minimizes the sharpness in the same way that SAM does, yielding better results on the ImageNet dataset with essentially the same computational cost as the base optimizer. Our codes are available at https://***/AngusDujw/SAF. Copyright © 2022, The Authors. All rights reserved.

关键词： Trajectories

Model Developmental Safety: A Retention-Centric Method and Applications in Vision-Language Models

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Li, Gang Yu, Wendi Yao, Yao Tong, Wei Liang, Yingbin Lin, Qihang Yang, Tianbao Department of Computer Science & Engineering Texas A&M University College Station United States Department of Mathematics The University of Iowa Iowa City United States Global Research and Development General Motors Warren United States Department of Electrical and Computer Engineering The Ohio State University Columbus United States Tippie College of Business The University of Iowa Iowa City United States

In the real world, a learning-enabled system usually undergoes multiple cycles of model development to enhance the system’s ability to handle difficult or emerging tasks, which involve collecting new data, training a new model and validating the model. This continual model development process raises a significant issue that the model development for acquiring new or improving existing capabilities may inadvertently lose capabilities of the old model, also known as catastrophic forgetting. Existing continual learning studies focus on mitigating catastrophic forgetting by trading off performance on previous tasks and new tasks to ensure good average performance. However, they are inadequate for many applications especially in safety-critical domains, as failure to strictly preserve the good performance of the old model not only introduces safety risks and uncertainties but also imposes substantial expenses in the re-improving and re-validation of existing properties. To address this issue, we introduce model developmental safety as a guarantee of a learning system such that in the model development process the new model should strictly preserve the existing protected capabilities of the old model while improving its performance on target tasks. To ensure the model developmental safety, we present a retention-centric framework by formulating the model developmental safety as data-dependent constraints. Under this framework, we study how to develop a pretrained vision-language model, specifically the CLIP model, for acquiring new capabilities or improving existing capabilities of image classification. We propose an efficient constrained optimization algorithm with theoretical guarantee and use its insights to finetune a CLIP model with task-dependent heads for promoting the model developmental safety. Our experiments on improving vision perception capabilities on autonomous driving and scene recognition datasets demonstrate the efficacy of the proposed approach.1 © 2024

关键词： Adversarial machine learning

Multiband highly sensitive metamaterial-based THz absorber for edible oil adulteration sensing

学校读者我要写书评

暂无评论

Optik 2024年 307卷

作者： Anowarul Haque, S.M. Sohayeb Hossain, Md Alqahtani, Abdulrahman Islam, Mohammad Tariqul Moniruzzaman, Md. Samsuzzaman, Md. Dept. of Electrical Electronic and Systems Engineering Faculty of Engineering and Built Environment Universiti Kebangsaan Malaysia Selangor Bangi43600 Malaysia Dept. of Electrical and Electronic Engineering Mymensingh Engineering College Bangladesh Department of Biomedical Technology College of Applied Medical Sciences in Al-Kharj Prince Sattam Bin Ab-dulaziz University Al-Kharj11942 Saudi Arabia Department of Medical Equipment Technology College of Applied Medical Science Majmaah University Majmaah City11952 Saudi Arabia Dept. of Electrical and Electronic Engineering College of Engineering and Technology International University of Business Agriculture and Technology Dhaka Uttara1230 Bangladesh Dept. of Computer and Communication Engineering Faculty of Computer Science and Engineering Patuakhali Science and Technology University Bangladesh Faculty of Graduate Studies Daffodil International University Daffodil Smart City Birulia Dhaka Savar1216 Bangladesh

A terahertz (THz) wave technology is very appealing for applications in detection in biomedicine, biological spectroscopy, material characterization, detection, imaging, sensing, and other convenient applications. Metamaterials (MMs) are artificial structures that have structures in a regular periodic arrangement and don't exist in nature. By utilizing a proposed terahertz metamaterial absorber, a broad spectrum of adulterated lubricants can be detected. It is an efficient detection process that enables the development of a THz detection approach by making use of the metamaterial as a sensor. The proposed terahertz absorber is analyzed and optimized using the CST Microwave Studio 2019 software. The article proposed a multiband THz metamaterial absorber with a unit size of 70 µm. According to the demonstrated results, we can achieve five narrow absorption peaks with perfect absorption within the frequency range between (0.2–4.2) Terahertz. Also, the result shows that, the average absorbance of 99.07% at the resonance frequencies of 0.56, 1.42, 2.676, 3.14, and 3.89 THz. The E (electric) and H (magnetic)-field distributions on the resonator at different resonance frequencies were also analyzed as well as. The recommended metamaterial absorber has a good sensing capability at different resonant frequencies, with refractive index sensitivities of 80 GHz/RIU, 240 GHz/RIU, 480 GHz/RIU, 480 GHz/RIU, and 520 GHz/RIU obtained by modifying the refractive index of the sample holder or the analytic layer, respectively. As a result, the proposed detector has a high refractive index sensitivity, high Q-factor, and high FOM, making it suitable for detecting oil adulteration. © 2024 Elsevier GmbH

关键词： Metamaterials

Continuous Image Outpainting with Neural ODE

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Gao, Penglei Yang, Xi Zhang, Rui Huang, Kaizhu Department of Computer Science University of Liverpool United Kingdom School of Advanced Technology Department of Foundational Mathematics Xi’an Jiaotong-Liverpool University China Institute of Applied Physical Sciences and Engineering Duke Kunshan University China

Different from image inpainting, generalised image outpainting aims to extend appealing content all-side around a given image. Existing state-of-the-art outpainting methods often rely on discrete extrapolation to extend the feature map in the bottleneck. They thus suffer from content unsmoothness, especially in circumstances where the outlines of objects in the extrapolated regions are incoherent with the input sub-images. To mitigate this issue, we design a novel bottleneck with Neural ODEs to make continuous extrapolation in latent space, which could be a plug-in for many deep learning frameworks. Our ODE-based network continuously transforms the state and makes accurate predictions by learning the incremental relationship among latent points, leading to both smooth and structured feature representation. Experimental results on three real-world datasets both applied on transformer-based and CNN-based frameworks show that our methods could generate more realistic and coherent images against the state-of-the-art image outpainting approaches. © 2023, The Authors. All rights reserved.

关键词： Extrapolation

Prediction of Passenger Car Friction Brake Application Based on Longitudinal Dynamics Using On-Road Vehicle Driving Data

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Jensen, Kenneth M. Santos, Ilmar F. Clemmensen, Line Katrine Harder Theodorsen, Søren Corstens, Harry J.P. Department of Civil and Mechanical Engineering Technical University of Denmark Kgs. Lyngby Denmark Department of Applied Mathematics and Computer Science Technical University of Denmark Kgs. Lyngby Denmark ILIAS Solutions Brussels Belgium

This work considers the problem of predicting the braking state of vehicles, i.e. when the friction brakes are being applied. The work is aimed at providing the foundation for the prediction of brake pad wear from vehicle usage data. The braking state is predicted from vehicle longitudinal dynamics, considering 6-axis inertial measurement unit (IMU), global navigation satellite system (GNSS), and vehicle on-board diagnostics (OBD-II) data, along with sensor fusion for more accurate vehicle state estimation. The proposed method aims at providing broad applicability to different vehicles, by using only standard OBD-II data instead of relying on manufacturer specific data. Rolling resistance coefficient and product of aerodynamic drag coefficient and frontal area are estimated using linear regression on data from vehicle coast-down tests with the drivetrain disengaged. Engine mass moment of inertia along with engine braking torque is estimated from similar tests with the drivetrain engaged in different gears. The engine braking torque is approximated using a polynomial expression in the engine speed. Braking state predictions are made by comparing measured longitudinal decelerations with estimated, assuming that if the measured exceeds the estimated, then the difference is caused by the application of the friction brakes. Experimental validation was carried out using two modern passenger cars. Comparisons of measured and estimated decelerations from coast-down tests with fitted parameters show good agreements when the drivetrain is disengaged, as well as engaged. For one vehicle labels for the braking state are obtained indirectly using a camera pointed at the brake pedal, using object tracking to determine if the brakes are being applied. For the other vehicle, the braking states are obtained more directly trough access to the brake pressure. Overall prediction performance is good with accuracy >94% and precision around 99%, while recall is around 90% across the two v

关键词： Parameter estimation