检索结果-内蒙古大学图书馆

High-performance UV polarization sensitive photodetector for a graphene(2D)/GaN(3D) junction with a non-centrosymmetric electric field

引用

Photonics Research 2025年第6期13卷 1544-1553页

作者： Can Zou Qing Liu Lu Zhang Xiao Tang Xiaohang Li Shuti Li Fangliang Gao Guangdong Engineering Research Center of Optoelectronic Functional Materials and Devices School of Electronic Science and Engineering (School of Microelectronics) South China Normal University Guangzhou 510631 China Advanced Semiconductor Laboratory Electrical and Computer Engineering Program CEMSE Division King Abdullah University of Science and Technology Thuwal 239556900 Saudi Arabia e-mail: lishuti@*** e-mail: gaofl@m.***

This study pioneers a high-performance UV polarization-sensitive photodetector by ingeniously integrating non-centrosymmetric metal nanostructures into a graphene ( Gr ) / Al 2 O 3 / GaN heterojunction. Unlike conventional approaches constrained by graphene’s intrinsic isotropy or complex nanoscale patterning, our design introduces asymmetric metal architectures (E-/T-type) to artificially create directional anisotropy. These structures generate plasmon-enhanced localized electric fields that selectively amplify photogenerated carrier momentum under polarized UV light (325 nm), synergized with Fowler-Nordheim tunneling (FNT) across an atomically thin Al 2 O 3 barrier. The result is a breakthrough in performance: a record anisotropy ratio of 115.5 (E-type, − 2 V ) and exceptional responsivity (97.7 A/W), surpassing existing graphene-based detectors by over an order of magnitude. Crucially, by systematically modulating metal geometry and density, we demonstrate a universal platform adaptable to diverse 2D/3D systems. This study provides a valuable reference for developing and practically applying photodetectors with higher anisotropy than ultraviolet polarization sensitivity.

关键词： Chemical vapor deposition Metals Nanostructures Quantum efficiency Surface plasmon resonance X-ray diffraction

来源：评论

学校读者我要写书评

暂无评论

Improved Model and Analysis for RIS-Assisted Indoor Terahertz Wireless Networks

arXiv

引用

arXiv 2024年

作者： Chai, Zhi Xu, Jiajie Alouini, Mohamed-Slim Coon, Justin P. The Department of Engineering Science University of Oxford Parks Road OxfordOX1 3PJ United Kingdom The Computer Electrical and Mathematical Sciences and Engineering Division King Abdullah University of Science and Technology Thuwal23955 Saudi Arabia

In this paper, we propose a new model for indoor THz communication assisted by RIS. We conduct a realistic modeling of indoor obstacles and analyze their impact on performance. Order statistics are applied to calculate the cumulative distribution functions (CDFs) of distances from the transmitter to the selected RIS, i.e., the nearest RIS in the bounded indoor environment to the transmitter, and from the selected RIS to the receiver. We calculate the coverage probability (CP) as a function of the RIS number, obstacle density, room size, and the transmitter’s location. By comparing the numerical results obtained from the analytical expressions with Monte Carlo simulations, we verify the accuracy of our analysis. Through the numerical results, it is observed that room size and obstacle density affect the CP in a significant way. However, by optimizing the transmitter’s location and increasing the RIS number deployed in the room, the CP can be significantly improved (e.g., an increase of around 15% by optimizing the transmitter’s location, and an increase of around 30% by increasing the RIS number deployed in the room). © 2024, CC BY.

关键词： Transmitters

来源：评论

学校读者我要写书评

暂无评论

A Robust Development of Calorie Prediction Methodology based on Artificial Intelligence Assisted Machine Learning Model

A Robust Development of Calorie Prediction Methodology based...

引用

Intelligent Technologies (CONIT), International Conference on

作者： G Naga Venkata RamaKrishna S Sreelakshmi Meenakshi Diwakar S. Ramkumar T Vinithra Banu T. Thilagam Faculty of Engineering University of Technology and Applied Sciences Al Musannah Sultanate of Oman Department of Computer Science and Engineering Saveetha Engineering College Chennai Tamil Nadu India Department of CSE (AI and ML) S.A.Engineering College Chennai Tamil Nadu India Department of Electronics and communication engineering Sri Eshwar College of Engineering Coimbatore Tamil Nadu India Department of Mechanical Engineering Prince Shri Venkateshwara Padmavathy Engineering College Chennai Tamil Nadu India Department of Computer Science & Engineering S.A Engineering college Chennai Tamil Nadu India

ISBN: (数字)9798350349900

ISBN: (纸本)9798350349917

Modern people’s hectic schedules are a direct result of the ways they live and the responsibilities they have at work. On the other hand, maintaining a healthy lifestyle demands consistent physical exertion. Obesity is a result of people not paying attention to what they eat. As a result of modern living, obesity is on the rise. As a result, in order to maintain a healthy weight, people pick their food and exercise habits accordingly. People should know how many calories they consume and how many calories they burn. Calorie counts are easy to get online or on product labels, so it’s not hard to keep track of one’s calorie intake. There aren’t many tools available to help you keep track of the calories you burn. To keep up with the hectic pace of modern life and the constant search for shortcuts, we built a system that uses machine learning to track post-workout calorie burn and development. By taking a few attributes as input, this system can approximate the calories burned, which will show daily growth and motivate people to exercise more. This paper presents an innovative ML model for accurate calorie prediction; it’s called the Improved Learning Model for Calorie Prediction (ILMCP). To test how well it works, the authors cross-validate it with the traditional Extreme Gradient Boosting (XGBoost) algorithm. As a result, it is the quickest approach and may be utilized for object identification in real-time. Next, we’ll apply the ILMCP algorithm to the picture for segmentation. Foreground extraction with little to no user input requires it. Following picture segmentation, we use the probing object’s known volume to determine the food item’s volume. Once the volume has been determined, the food item’s mass may be determined using formulae. Then, the connection between weight and calories can be used to compute the food item’s calories.

关键词： Obesity Schedules Accuracy Target recognition Computational modeling Predictive models Feature extraction

来源：评论

学校读者我要写书评

暂无评论

IOT and Sensor Based Technique to Track and Monitor the Work of the Street Sweepers 4

IOT and Sensor Based Technique to Track and Monitor the Work...

引用

4th International Conference on Smart Electronics and Communication, ICOSEC 2023

作者： Jayaudhaya, J. Srinivas, Amgoth Srinivasan, A. Vijayalakshmi, G. Karthikeyan, M. Kumar, A. R. M. D. Engineering College Department of Electronics and Communication Engineering Tamil Nadu Chennai India Narasimha Reddy Engineering College Department of Electronics and Communication Engineering Maisammaguda Telangana India Dhanalakshmi College of Engineering Department of Ai and Ds Tamil Nadu Chennai India Prince Shri Venkateshwara Padmavathy Engineering College Department of Electronics and Communication Engineering Tamil Nadu Chennai India R.M.K. College of Engineering and Technology Department of Computer Science Engineering Tamil Nadu Chennai India Dhanalakshmi College of Engineering Department of Electronics and Communication Engineering Tamil Nadu Chennai India

ISBN: (纸本)9798350300888

To inform the public about environmental issues, scientists are developing portable sensors. However, data collection alone is insufficient for effective environmental action. This study emphasizes the proposed innovative environmental sensing technology, which is currently being utilized on street sweepers in an American metropolis. The proposed product was informed by industry executives' requirements and refined as an environmental action project through consultation. This article examines the environmental action landscape qualitatively to better equip innovators to design efficient technological solutions. Current environmental initiatives, the environmental effects of street cleansing, and related research on environmental sensing technologies are examined. The proposed research has implications for design fieldwork, sensor system performance evaluation, and ecological consciousness. Proposed innovation paves the path for environmental action by bridging the divide between observation and response. © 2023 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Financial Opinion Mining

Financial Opinion Mining

引用

2021 Conference on Empirical Methods in Natural Language Processing, EMNLP 2021

作者： Chen, Chung-Chi Huang, Hen-Hsen Chen, Hsin-Hsi Department of Computer Science and Information Engineering National Taiwan University Taiwan Institute of Information Science Academia Sinica Taiwan MOST Joint Research Center for AI Technology and All Vista Healthcare Taiwan

来源：评论

学校读者我要写书评

暂无评论

Exploring Iterative Enhancement for Improving Learnersourced Multiple-Choice Question Explanations with Large Language Models

arXiv

引用

arXiv 2023年

作者： Bao, Qiming Leinonen, Juho Peng, Alex Yuxuan Zhong, Wanjun Gendron, Gaël Pistotti, Tim Huang, Alice Denny, Paul Witbrock, Michael Liu, Jiamou Strong AI Lab NAOInstitute Waipapa Taumata Rau The University of Auckland New Zealand Xtracta New Zealand School of Computer Science University of Auckland New Zealand School of Computer Science and Engineering Sun Yat-Sen University China School of Life and Environmental Sciences University of Sydney Australia

Large language models (LLMs) have demonstrated strong capabilities in language understanding and generation, and their potential in educational contexts is increasingly being explored. One promising area is learnersourcing, where students engage in creating their own educational content, such as multiple-choice questions. A critical step in this process is generating effective explanations for the solutions to these questions, as such explanations aid in peer understanding and promote deeper conceptual learning. However, students often find it difficult to craft high-quality explanations due to limited understanding or gaps in their subject knowledge. To support this task, we introduce "ILearner-LLM," a framework that uses iterative enhancement with LLMs to improve generated explanations. The framework combines an explanation generation model and an explanation evaluation model fine-tuned using student preferences for quality, where feedback from the evaluation model is fed back into the generation model to refine the output. Our experiments with LLaMA2-13B and GPT-4 using five large datasets from the PeerWise MCQ platform show that ILearner-LLM produces explanations of higher quality that closely align with those written by students. Our findings represent a promising approach for enriching the learnersourcing experience for students and for leveraging the capabilities of large language models for educational applications. © 2023, CC BY-NC-ND.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

APPROXIMATION RATES FOR SHALLOW RELUk NEURAL NETWORKS ON SOBOLEV SPACES VIA THE RADON TRANSFORM

arXiv

引用

arXiv 2024年

作者： Mao, Tong Siegel, Jonathan W. Xu, Jinchao Computer Electrical and Mathematical Science and Engineering Division King Abdullah University of Science and Technology Thuwal23955 Saudi Arabia Department of Mathematics Texas A&M University College StationTX77843 United States

Let Ω ⊂ d be a bounded domain. We consider the problem of how efficiently shallow neural networks with the ReLUk activation function can approximate functions from Sobolev spaces W s(Lp(Ω)) with error measured in the Lq(Ω)-norm. Utilizing the Radon transform and recent results from discrepancy theory, we provide a simple proof of nearly optimal approximation rates in a variety of cases, including when q ≤ p, p ≥ 2, and s ≤ k + (d + 1)/2. The rates we derive are optimal up to logarithmic factors, and significantly generalize existing results. An interesting consequence is that the adaptivity of shallow ReLUk neural networks enables them to obtain optimal approximation rates for smoothness up to order s = k + (d + 1)/2, even though they represent piecewise polynomials of fixed degree *** Codes 62M45, 41A25, 41A30 Copyright © 2024, The Authors. All rights reserved.

关键词： Sobolev spaces

来源：评论

学校读者我要写书评

暂无评论

Predictability performance enhancement for suspended sediment in rivers: Inspection of newly developed hybrid adaptive neuro-fuzzy system model

引用

International Journal of Sediment Research 2022年第3期37卷 383-398页

作者： Rana Muhammad Adnan Zaher Mundher Yaseen Salim Heddam Shamsuddin Shahid Aboalghasem Sadeghi-Niaraki Ozgur Kisi State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering Hohai University Department of Urban Planning Engineering Networks and Systems Institute of Architecture and Construction South Ural State University New Era and Development in Civil Engineering Research Group Scientific Research Center Al-Ayen University Faculty of Science Agronomy Department Hydraulics Division University School of Civil Engineering Faculty of Engineering Universiti Teknologi Malaysia (UTM) Geoinformation Tech.Center of Excellence Faculty of Geomatics Engineering K.N.Toosi University of Technology Department of Computer Science and Engineering Sejong University Civil Engineering Department Ilia State University

Reliable modeling of river sediments transport is important as it is a defining factor of the economic viability of dams, the durability of hydroelectric-equipment, river susceptibility to pollution, suitability for navigation, and potential for aesthetics and fish habitat. The capability of a new machine learning model, fuzzy c-means based neuro-fuzzy system calibrated using the hybrid particle swarm optimization-gravitational search algorithm（ANFIS-FCM-PSOGSA） in improving the estimation accuracy of river suspended sediment loads（SSLs） is investigated in the current study. The outcomes of the proposed method were compared with those obtained using the fuzzy c-means based neuro-fuzzy system calibrated using particle swarm optimization（ANFIS-FCM-PSO）, ANFIS-FCM, and sediment rating curve（SRC） models. Various input combinations involving lagged river flow（Q） and suspended sediment（S） values were used for model development. The effect of Q and S on the model's accuracy also was assessed by including the difference between lagged Q and S values as inputs. The model performance was assessed using the root mean square error（RMSE）, mean absolute error（MAE）, Nash-Sutcliffe Efficiency（NSE）, and coefficient of determination（R2） and several graphical comparison methods. The results showed that the proposed model enhanced the prediction performance of the ANFIS-FCM-PSO（or ANFIS-FCM） models by 8.14%（1.72%）, 14.7%（5.71%）, 12.5%（2.27%）, and 25.6%（1.86%）,in terms of the RMSE, MAE, NSE and R2, respectively. The current study established the potential of the proposed ANFIS-FCM-PSOGSA model for simulation of the cumulative sediment load. The modeling results revealed the potential effects of the river flow lags on the sediment transport quantification.

关键词： Suspended sediment load Adaptive neuro-fuzzy system Particle swarm optimization Gravitational search algorithm

来源：评论

学校读者我要写书评

暂无评论

HOW BENIGN IS BENIGN OVERFITTING? 9

HOW BENIGN IS BENIGN OVERFITTING?

引用

9th International Conference on Learning Representations, ICLR 2021

作者： Sanyal, Amartya Kanade, Varun Dokania, Puneet K. Torr, Philip H.S. Department of Computer Science University of Oxford Oxford United Kingdom The Alan Turing Institute London United Kingdom Department of Engineering Science University of Oxford Oxford United Kingdom Five AI Limited

We investigate two causes for adversarial vulnerability in deep neural networks: bad data and (poorly) trained models. When trained with SGD, deep neural networks essentially achieve zero training error, even in the presence of label noise, while also exhibiting good generalization on natural test data, something referred to as benign overfitting (Bartlett et al., 2020;Chatterji & Long, 2020). However, these models are vulnerable to adversarial attacks. We identify label noise as one of the causes for adversarial vulnerability, and provide theoretical and empirical evidence in support of this. Surprisingly, we find several instances of label noise in datasets such as MNIST and CIFAR, and that robustly trained models incur training error on some of these, i.e. they don't fit the noise. However, removing noisy labels alone does not suffice to achieve adversarial robustness. We conjecture that in part sub-optimal representation learning is also responsible for adversarial vulnerability. By means of simple theoretical setups, we show how the choice of representation can drastically affect adversarial robustness. © 2021 ICLR 2021 - 9th International Conference on Learning Representations. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Referenceless User Controllable Semantic Image Synthesis

Referenceless User Controllable Semantic Image Synthesis

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Jonghyun Kim Gen Li Joongkyu Kim AI Lab CTO Division LG Electronics Seoul Republic of Korea School of Informatics University of Edinburgh Edinburgh UK Dept. of Electrical and Computer Engineering Sungkyunkwan University Suwon Republic of Korea

Despite recent progress in semantic image synthesis, complete control over image style remains a challenging problem. Existing methods require reference images to feed style information into semantic layouts, which indicates that the style is constrained by the given image. In this paper, we propose a model named RUCGAN for user controllable semantic image synthesis, which utilizes a singular color to represent the style of a specific semantic region. The proposed network achieves reference-free semantic image synthesis by injecting color as user-desired styles into each semantic layout, and is able to synthesize semantic images with unusual colors. Extensive experimental results on various challenging datasets show that the proposed method outperforms existing methods, and we further provide an interactive UI to demonstrate the advantage of our approach for style controllability. The codes and UI are available at: https://***/BenjaminJonghyun/RUCGAN

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：