检索结果-内蒙古大学图书馆

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

Trident-shaped Dual Band Monopole Antenna with Defected Ground Plane for 5G Applications

学校读者我要写书评

暂无评论

Trident-shaped Dual Band Monopole Antenna with Defected Grou...

2022 IEEE Wireless Antenna and Microwave Symposium, WAMS 2022

作者： Sharma, Preeti Tiwari, Rakesh N. Singh, Prabhakar Kumar, Pradeep School of Basic and Applied Sciences Galgotias University Physics Division Greater Noida India Greater Noida Institute of Technology Department of Applied Science Greater Noida India School of Computational and Integrative Sciences Jawaharlal Nehru University New Delhi India Electronic and Computer Engineering University of KwaZulu-Natal Discipline of Electrical Durban South Africa

ISBN: (数字)9781665458467

ISBN: (纸本)9781665458467

This work explores a compact dual band trident-shaped monopole antenna for 5G applications. The designed antenna has dimension 32 × 25 × 1.524 mm3. The structure of the antenna consists of symmetric trident-shaped patch with arrow shaped strip in the center. The L-shaped slots are used to modify the defected ground plane and to obtain the dual band characteristics. The proposed design demonstrates the measured impedance bandwidths in the range 3.27-3.88 GHz (lower band) and 4.85-5.23 GHz (upper band), respectively. The radiation efficiency for both the operating bands are above 76%, the realized average gains are 1.28 dBi and 1.7 dBi across both the operating bands, respectively. The radiation patterns of the monopole antenna are acceptable for 5G applications. The proposed design is modeled and simulated using CST microwave studio. The simulated results agreeing well with the fabricated antenna. © 2022 IEEE.

关键词： Monopole antennas

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

Unbiased Estimation using Underdamped Langevin Dynamics

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Ruzayqat, Hamza Chada, Neil K. Jasra, Ajay Applied Mathematics and Computational Science Program Computer Electrical and Mathematical Sciences and Engineering Division King Abdullah University of Science and Technology Thuwal23955-6900 Saudi Arabia

In this work we consider the unbiased estimation of expectations w.r.t. probability measures that have non-negative Lebesgue density, and which are known point-wise up-to a normalizing constant. We focus upon developing an unbiased method via the underdamped Langevin dynamics, which has proven to be popular of late due to applications in statistics and machine learning. Specifically in continuous-time, the dynamics can be constructed so that as the time goes to infinity they admit the probability of interest as a stationary measure. In many cases, time-discretized versions of the underdamped Langevin dynamics are used in practice which are run only with a fixed number of iterations. We develop a novel scheme based upon doubly randomized estimation as in [17, 19], which requires access only to time-discretized versions of the dynamics. The proposed scheme aims to remove the dicretization bias and the bias resulting from running the dynamics for a finite number of iterations. We prove, under standard assumptions, that our estimator is of finite variance and either has finite expected cost, or has finite cost with a high probability. To illustrate our theoretical findings we provide numerical experiments which verify our theory, which include challenging examples from Bayesian statistics and statistical *** Codes 60J22, 65C05, 65C40, 82C31, 62G08, 35Q56 © 2022, CC BY.

关键词： Dynamics

Data Assimilation for Sign-indefinite Priors: A generalization of Sinkhorn's algorithm

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Dong, Anqi Georgiou, Tryphon T. Tannenbaum, Allen Department of Mechanical and Aerospace Engineering University of California IrvineCA92697 United States Department of Computer Science and Applied Mathematics & Statistics The State University of New York Stony BrookNY11794 United States

The purpose of this work is to develop a framework to calibrate signed datasets so as to be consistent with specified marginals by suitably extending the Schrödinger-Fortet-Sinkhorn paradigm. Specifically, we seek to revise sign-indefinite multi-dimensional arrays in a way that the updated values agree with specified marginals. Our approach follows the rationale in Schrödinger's problem [25, 26], aimed at updating a "prior" probability measure to agree with marginal distributions. The celebrated Sinkhorn's algorithm [28, 18] (established earlier by R. Fortet [13, 11]) that solves Schrödinger's problem found early applications in calibrating contingency tables in statistics and, more recently, multi-marginal problems in machine learning and optimal transport. Herein, we postulate a sign-indefinite prior in the form of a multi-dimensional array, and propose an optimization problem to suitably update this prior to ensure consistency with given marginals. The resulting algorithm generalizes the Sinkhorn algorithm in that it amounts to iterative scaling of the entries of the array along different coordinate directions. The scaling is multiplicative but also, in contrast to Sinkhorn, inverse-multiplicative depending on the sign of the entries. Our algorithm reduces to the classical Sinkhorn algorithm when the entries of the prior are *** Codes 49M29, 90C25 Copyright © 2023, The Authors. All rights reserved.

关键词： Iterative methods

Large regenerative parametric amplification on chip at ultra-low pump powers

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Zhao, Yun Jang, Jae K. Ji, Xingchen Okawachi, Yoshitomo Lipson, Michal Gaeta, Alexander L. Department of Applied Physics and Applied Mathematics Columbia University New YorkNY10027 United States Department of Electrical Engineering Columbia University New YorkNY10027 United States

Chip-based optical amplifiers can significantly expand the functionalities of photonic devices. In particular, optical-parametric amplifiers (OPAs), with engineerable gain-spectra, are well-suited for nonlinear-photonic applications. Chip-based OPAs typically require long waveguides that occupy a large footprint, and high pump powers that cannot be easily produced with chip-scale lasers. We theoretically and experimentally demonstrate a microresonator-assisted regenerative OPA that benefits from the large nonlinearity enhancement of microresonators and yields a high gain in a small footprint. We achieve 30-dB parametric gain with only 9 mW of cw-pump power and show that the gain spectrum can be engineered to cover telecom channels inaccessible with Er-based amplifiers. We further demonstrate the amplification of Kerr-soliton comb lines and the preservation of their phase properties. Additionally, we demonstrate amplification by injection locking of optical-parametric oscillators, which corresponds to a regenerative amplifier pumped above the oscillation threshold. Novel dispersion engineering techniques such as coupled cavities and higher-order-dispersion phase matching can further extend the tunability and spectral coverage of our amplification schemes. The combination of high gain, small footprint, low pump power, and flexible gain-spectra engineering of our regenerative OPA is ideal for amplifying signals from the nanowatt to microwatt regimes for portable or space-based devices where ultralow electrical power levels are required and can lead to important applications in on-chip optical- and microwave-frequency synthesis and precise timekeeping. © 2022, CC BY-NC-ND.

关键词： Optical parametric amplifiers

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

Beamforming in Integrated Sensing and Communication Systems with Reconfigurable Intelligent Surfaces

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Sankar, R.S. Prasobh Chepuri, Sundeep Prabhakar Eldar, Yonina C. The Department of Electrical Communication Engineering Indian Institute of Science Bangalore India The Department of Mathematics and Computer Science Weizmann Institute of Science Rehovot Israel

We consider transmit beamforming and reflection pattern design in reconfigurable intelligent surface (RIS)-assisted integrated sensing and communication (ISAC) systems to jointly precode communication symbols and radar waveforms. We treat two settings of multiple users and targets. In the first, we use a single RIS to enhance the communication performance of the ISAC system and design beams with good cross-correlation properties to match a desired beam pattern while guaranteeing a desired signal-to-interference plus noise ratio (SINR) for each user. In the second setting, we use two dedicated RISs to aid the ISAC system, wherein the beams are designed to maximize the worst-case target illumination power while guaranteeing a desired SINR for each user. We propose solvers based on alternating optimization as the design problems in both cases are non-convex optimization problems. Through a number of numerical simulations, we demonstrate the advantages of RIS-assisted ISAC systems. In particular, we show that the proposed single-RIS assisted ISAC system improves the minimum user SINR while suffering from a moderate loss in radar target illumination power. On the other hand, the dual-RIS assisted ISAC system improves both minimum user SINR as well as worst-case target illumination power at the targets, especially when the users and targets are not directly visible. Copyright © 2022, The Authors. All rights reserved.

关键词： Beamforming