检索结果-内蒙古大学图书馆

Universal linguistic inductive biases via meta-learning 42

学校读者我要写书评

暂无评论

Universal linguistic inductive biases via meta-learning

42nd Annual Meeting of the Cognitive science Society: Developing a Mind: Learning in Humans, Animals, and Machines, CogSci 2020

作者： McCoy, R. Thomas Grant, Erin Smolensky, Paul Griffiths, Thomas L. Linzen, Tal Department of Cognitive Science Johns Hopkins University United States Department of Electrical Engineering & Computer Sciences University of California Berkeley United States Microsoft Research AI RedmondWA United States Departments of Psychology and Computer Science Princeton University United States

How do learners acquire languages from the limited data available to them? This process must involve some inductive biases-factors that affect how a learner generalizes-but it is unclear which inductive biases can explain observed patterns in language acquisition. To facilitate computational modeling aimed at addressing this question, we introduce a framework for giving particular linguistic inductive biases to a neural network model;such a model can then be used to empirically explore the effects of those inductive biases. This framework disentangles universal inductive biases, which are encoded in the initial values of a neural network's parameters, from non-universal factors, which the neural network must learn from data in a given language. The initial state that encodes the inductive biases is found with meta-learning, a technique through which a model discovers how to acquire new languages more easily via exposure to many possible languages. By controlling the properties of the languages that are used during meta-learning, we can control the inductive biases that meta-learning imparts. We demonstrate this framework with a case study based on syllable structure. First, we specify the inductive biases that we intend to give our model, and then we translate those inductive biases into a space of languages from which a model can meta-learn. Finally, using existing analysis techniques, we verify that our approach has imparted the linguistic inductive biases that it was intended to impart. © 2020 The Author(s)

关键词： Linguistics

Image super-resolution with an enhanced group convolutional neural network

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Tian, Chunwei Yuan, Yixuan Zhang, Shichao Lin, Chia-Wen Zuo, Wangmeng Zhang, David School of Software Northwestern Polytechnical University Shaanxi Xi’an710129 China National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology Shaanxi Xi’an710129 China Department of Electrical Engineering City University of Hong Kong Hong Kong School of Computer Science and Engineering Central South University Hunan Changsha410083 China Department of Electrical Engineering the Institute of Communications Engineering National Tsing Hua University Hsinchu Taiwan School of Computer Science and Technology Harbin Institute of Technology Heilongjiang Harbin150001 China Peng Cheng Laboratory Guangdong Shenzhen518055 China Guangdong Shenzhen518172 China Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen China

CNNs with strong learning abilities are widely chosen to resolve super-resolution problem. However, CNNs depend on deeper network architectures to improve performance of image super-resolution, which may increase computational cost in general. In this paper, we present an enhanced super-resolution group CNN (ESRGCNN) with a shallow architecture by fully fusing deep and wide channel features to extract more accurate low-frequency information in terms of correlations of different channels in single image super-resolution (SISR). Also, a signal enhancement operation in the ESRGCNN is useful to inherit more long-distance contextual information for resolving long-term dependency. An adaptive up-sampling operation is gathered into a CNN to obtain an image super-resolution model with low-resolution images of different sizes. Extensive experiments report that our ESRGCNN surpasses the state-of-the-arts in terms of SISR performance, complexity, execution speed, image quality evaluation and visual effect in SISR. Code is found at https://***/hellloxiaotian/ESRGCNN. Copyright © 2022, The Authors. All rights reserved.

关键词： Convolution

How Social Robots Influence People’s Trust in Critical Situations

学校读者我要写书评

暂无评论

How Social Robots Influence People’s Trust in Critical Situ...

IEEE International Workshop on Robot and Human Communication (ROMAN)

作者： Alessandra Rossi Kerstin Dautenhahn Kheng Lee Koay Michael L. Walters School of Engineering and Computer Science University of Hertfordshire United Kingdom Departments of Electrical and Computer Engineering/Systems Design Engineering University of Waterloo Canada

ISBN: (数字)9781728160757

ISBN: (纸本)9781728160764

As we expect that the presence of autonomous robots in our everyday life will increase, we must consider that people will have not only to accept robots to be a fundamental part of their lives, but they will also have to trust them to reliably and securely engage them in collaborative tasks. Several studies showed that robots are more comfortable interacting with robots that respect social conventions. However, it is still not clear if a robot that expresses social conventions will gain more favourably people's trust. In this study, we aimed to assess whether the use of social behaviours and natural communications can affect humans' sense of trust and companionship towards the robots. We conducted a between-subjects study where participants' trust was tested in three scenarios with increasing trust criticality (low, medium, high) in which they interacted either with a social or a non-social robot. Our findings showed that participants trusted equally a social and non-social robot in the low and medium consequences scenario. On the contrary, we observed that participants' choices of trusting the robot in a higher sensitive task was affected more by a robot that expressed social cues with a consequent decrease of their trust in the robot.

关键词：

Floquet vortex states induced by light carrying an orbital angular momentum

学校读者我要写书评

暂无评论

Physical Review B 2022年第8期105卷 L081301-L081301页

作者： Hwanmun Kim Hossein Dehghani Iman Ahmadabadi Ivar Martin Mohammad Hafezi Joint Quantum Institute NIST and University of Maryland College Park Maryland 20742 USA Department of Physics University of Maryland College Park Maryland 20742 USA Departments of Electrical and Computer Engineering and Institute for Research in Electronics and Applied Physics University of Maryland College Park Maryland 20742 USA Materials Science Division Argonne National Laboratory Argonne Illinois 60439 USA

We propose a scheme to create an electronic Floquet vortex state by irradiating a two-dimensional semiconductor with a laser light carrying nonzero orbital angular momentum. We analytically and numerically study the properties of the Floquet vortex states, with methods analogous to the ones previously applied to the analysis of superconducting vortex states. We show that such Floquet vortex states are similar to superconducting vortex states, and they exhibit a wide range of tunability. To illustrate the potential utility of such tunability, we show how such states could be used for quantum state engineering.

关键词： Beam optics Optical vortices Photoinduced effect Semiconductor quantum optics 2-dimensional systems Floquet systems Quantum wells Semiconductors Optical absorption spectroscopy

Superior Performance of ZnGaO Solar-Blind Photodetectors by Implementing TFT Structure and Tunable ZNO Cycle Ratio

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Fan, Hui-Chen Wang, Chen Xu, Yi-Hong Fan, Teng-Min Huang, Pao-Hsun Ruan, Yu-Jiao Wu, Tingzhu Kuo, Hao-Chung Wuu, Dong-Sing Lai, Feng-Min Lien, Shui-Yang Xiamen Key Laboratory of Development and Application for Advanced Semiconductor Coating Technology School of Opto-Electronic and Communication Engineering Xiamen University of Technology Xiamen361024 China School of Ocean Information Engineering Jimei University Jimei District Fujian Xiamen361021 China National Measurement and Testing Center for Flat Panel Display Industry Xiamen Institute of Measurement and Testing Xiamen361024 China National Integrated Circuit Industry and Education Integration Innovation Platform Department of Electronic Science Xiamen University Xiamen361005 China Department of Photonics Graduate Institute of Electro-Optical Engineering College of Electrical and Computer Engineering National Yang Ming Chiao Tung University Hsinchu30010 Taiwan Department of Applied Materials and Optoelectronic Engineering National Chi Nan University Nantou54561 Taiwan Department of Biomedical Engineering Da-Yeh University Changhua51591 Taiwan

In this study, plasma enhanced atomic layer deposition was used to prepare ZnGaO thin films with different ZnO cycle ratios from 0% to 50%, which were employed to fabricate the MSM-type and TFT-type solar-blind photodetector. The properties of the films and the performance of the solar blind photodetector are greatly affected by the ZnO cycle ratio. Appropriate Zn doping content decreases the ratio of oxygen vacancies and Zn interstitials in Ga2O3, while excessive Zn doping degrades the quality of the film. Compared to MSM-type photodetector, an obviously improved performance of ZnGaO TFT-type photodetector has been observed due to its self-amplified photocurrent and the reduced dark current controlled by gate voltage. At a ZnO cycle ratio of 40%, a superior performance of TFT-type solar-blind photodetector with an ultra-low dark current ~4.5×10-13 A, an excellent light switching ratio ~2.2×107, and an outstandingly high sensitivity ~34.31 A/W has been obtained. These results demonstrate that the incorporation of Zn dopants can effectively modulate the performance of Ga2O3 TFT-type photodetector providing a great perspective for cost-effective and high-performance solar-blind PDs. © 2024, The Authors. All rights reserved.

关键词： Zinc oxide

Quantum Fourier Transform using Dynamic Circuits

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Bäumer, Elisa Tripathi, Vinay Seif, Alireza Lidar, Daniel Wang, Derek S. IBM Quantum IBM Research – Zurich Rüschlikon8803 Switzerland Department of Physics & Astronomy Center for Quantum Information Science & Technology University of Southern California Los AngelesCA90089 United States IBM Quantum IBM T.J. Watson Research Center Yorktown HeightsNY10598 United States Departments of Electrical & Computer Engineering Chemistry Physics & Astronomy Center for Quantum Information Science & Technology University of Southern California Los AngelesCA90089 United States

In dynamic quantum circuits, classical information from mid-circuit measurements is fed forward during circuit execution. This emerging capability of quantum computers confers numerous advantages that can enable more efficient and powerful protocols by drastically reducing the resource requirements for certain core algorithmic primitives. In particular, in the case of the n-qubit quantum Fourier transform followed immediately by measurement, the scaling of resource requirements is reduced from O(n2) two-qubit gates in an all-to-all connectivity in the standard unitary formulation to O(n) mid-circuit measurements in its dynamic counterpart without any connectivity constraints. Here, we demonstrate the advantage of dynamic quantum circuits for the quantum Fourier transform on IBM’s superconducting quantum hardware with certified process fidelities of > 50% on up to 16 qubits and > 1% on up to 37 qubits, exceeding previous reports across all quantum computing platforms. These results are enabled by our contribution of an efficient method for certifying the process fidelity, as well as of a dynamical decoupling protocol for error suppression during mid-circuit measurements and feed-forward within a dynamic quantum circuit. Our results demonstrate the advantages of leveraging dynamic circuits in optimizing the compilation of quantum algorithms. © 2024, CC BY.

关键词： Timing circuits

Growth and characterization of semipolar InGaN/GaN multiple quantum wells and light-emitting diodes on % MathType!MTEF!2!1!+- % feaagKart1ev2aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn % hiov2DGi1BTfMBaeXatLxBI9gBaerbd9wDYLwzYbItLDharqqr1ngB % PrgifHhDYfgasaacH8srps0lbbf9q8WrFfeuY-Hhbbf9v8qqaqFr0x % c9pk0xbba9q8WqFfea0-yr0RYxir-Jbba9q8aq0-yq-He9q8qqQ8fr % Fve9Fve9Ff0dmeaabaqaciGacaGaaeqabaWaaeaaeaaakeaadaqada % qaaiaaigdacaaIWaWaa0aaaeaacaaIXaaaamaanaaabaGaaGymaaaa % aiaawIcacaGLPaaaaaa!3C41! GaN templates

学校读者我要写书评

暂无评论

MRS Online Proceedings Library 2021年第1期892卷 127-132页

作者： Arpan Chakraborty T. Onuma T. J. Baker S. Keller S. F. Chichibu S. P. DenBaars S. Nakamura J. S. Speck U. K. Mishra Electrical and Computer Engineering and Materials Departments University of California Santa Barbara USA NICP/ERATO Japan Science and Technology Agency Japan Institute of Applied Physics and Graduate School of Pure and Applied Sciences University of Tsukuba Tsukuba Japan

InGaN/GaN MQW samples were grown by metal organic chemical vapor deposition (MOCVD) on % MathType!MTEF!2!1!+- % feaagKart1ev2aaatCvAUfeBSjuyZL2yd9gzLbv... 详细信息

InGaN/GaN MQW samples were grown by metal organic chemical vapor deposition (MOCVD) on % MathType!MTEF!2!1!+- % feaagKart1ev2aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn % hiov2DGi1BTfMBaeXatLxBI9gBaerbd9wDYLwzYbItLDharqqr1ngB % PrgifHhDYfgasaacH8srps0lbbf9q8WrFfeuY-Hhbbf9v8qqaqFr0x % c9pk0xbba9q8WqFfea0-yr0RYxir-Jbba9q8aq0-yq-He9q8qqQ8fr % Fve9Fve9Ff0dmeaabaqaciGacaGaaeqabaWaaeaaeaaakeaadaqada % qaaiaaigdacaaIWaWaa0aaaeaacaaIXaaaamaanaaabaGaaGymaaaa % aiaawIcacaGLPaaaaaa!3C41! $$\left( {10\overline 1 \overline 1 } \right)$$ oriented GaN templates. Effects of growth temperature and reactor pressure on the photoluminescence (PL) properties were investigated. The emission intensity improved significantly when the QWs were grown at 100 Torr, compared to higher pressure growths. The effect of well-width on the luminescence properties was investigated and an optimum well width of 40 Å was determined. Excitation dependent PL measurements revealed no shift in the PL emission wavelength suggesting the absence of electric field in the quantum wells. Furthermore, LEDs fabricated on % MathType!MTEF!2!1!+- % feaagKart1ev2aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn % hiov2DGi1BTfMBaeXatLxBI9gBaerbd9wDYLwzYbItLDharqqr1ngB % PrgifHhDYfgasaacH8srps0lbbf9q8WrFfeuY-Hhbbf9v8qqaqFr0x % c9pk0xbba9q8WqFfea0-yr0RYxir-Jbba9q8aq0-yq-He9q8qqQ8fr % Fve9Fve9Ff0dmeaabaqaciGacaGaaeqabaWaaeaaeaaakeaadaqada % qaaiaaigdacaaIWaWaa0aaaeaacaaIXaaaamaanaaabaGaaGymaaaa % aiaawIcacaGLPaaaaaa!3C41! $$\left( {10\overline 1 \overline 1 } \right)$$ GaN templates, emitting at 439 nm, showed no shift in the EL emission wavelength with the increase in drive current, reconfirming the absence of polarization.

关键词： nitride metalorganic deposition optoelectronic

Valley-polarized Exitonic Mott Insulator in WS2/WSe2 Moiré Superlattice

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Lian, Zhen Meng, Yuze Ma, Lei Maity, Indrajit Yan, Li Wu, Qiran Huang, Xiong Chen, Dongxue Chen, Xiaotong Chen, Xinyue Blei, Mark Taniguchi, Takashi Watanabe, Kenji Tongay, Sefaattin Lischner, Johannes Cui, Yong-Tao Shi, Su-Fei Department of Chemical and Biological Engineering Rensselaer Polytechnic Institute TroyNY12180 United States Departments of Materials The Thomas Young Centre for Theory and Simulation of Materials Imperial College London United Kingdom Department of Physics and Astronomy University of California RiversideCA92521 United States Department of Materials Science and Engineering University of California RiversideCA92521 United States Materials Science and Engineering Program School for Engineering of Matter Transport and Energy Arizona State University TempeAZ85287 United States International Center for Materials Nanoarchitectonics National Institute for Materials Science 1-1 Namiki Tsukuba305-0044 Japan Research Center for Functional Materials National Institute for Materials Science 1-1 Namiki Tsukuba305-0044 Japan Department of Electrical Computer & Systems Engineering Rensselaer Polytechnic Institute TroyNY12180 United States

Strongly enhanced electron-electron interaction in semiconducting moiré superlattices formed by transition metal dichalcogenides (TMDCs) heterobilayers has led to a plethora of intriguing fermionic correlated states1–7. Meanwhile, interlayer excitons in a type-II aligned TMDC heterobilayer moiré superlattice, with electrons and holes separated in different layers8,9, inherit this enhanced interaction and strongly interact with each other, promising for realizing tunable correlated bosonic quasiparticles10,11 with valley degree of freedom. We employ photoluminescence spectroscopy to investigate the strong repulsion between interlayer excitons and correlated electrons in a WS2/WSe2 moiré superlattice and combine with theoretical calculations to reveal the spatial extent of interlayer excitons and the band hierarchy of correlated states. We further find that an excitonic Mott insulator state emerges when one interlayer exciton occupies one moiré cell, evidenced by emerging photoluminescence peaks under increased optical excitation power. Double occupancy of excitons in one unit cell requires overcoming the energy cost of exciton-exciton repulsion of about 30-40 meV, depending on the stacking configuration of the WS2/WSe2 heterobilayer. Further, the valley polarization of the excitonic Mott insulator state is enhanced by nearly one order of magnitude. Our study demonstrates the WS2/WSe2 moiré superlattice as a promising platform for engineering and exploring new correlated states of fermion, bosons, and a mixture of both12 © 2023, CC BY.

关键词： Excitons

A heterogeneous group CNN for image super-resolution

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Tian, Chunwei Zhang, Yanning Zuo, Wangmeng Lin, Chia-Wen Zhang, David Yuan, Yixuan The School of Software Northwestern Polytechnical University Shaanxi Xi’an710129 China The National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology Shaanxi Xi’an710129 China The School of Computer Science Northwestern Polytechnical University The National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology Shaanxi Xi’an710129 China The School of Computer Science and Technology Harbin Institute of Technology Heilongjiang Harbin150001 China The Department of Electrical Engineering The Institute of Communications Engineering National Tsing Hua University Hsinchu Taiwan Guangdong Shenzhen518172 China The Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen China The Department of Electronic Engineering The Chinese University of Hong Kong Hong Kong

Convolutional neural networks (CNNs) have obtained remarkable performance via deep architectures. However, these CNNs often achieve poor robustness for image superresolution (SR) under complex scenes. In this paper, we present a heterogeneous group SR CNN (HGSRCNN) via leveraging structure information of different types to obtain a high-quality image. Specifically, each heterogeneous group block (HGB) of HGSRCNN uses a heterogeneous architecture containing a symmetric group convolutional block and a complementary convolutional block in a parallel way to enhance internal and external relations of different channels for facilitating richer low-frequency structure information of different types. To prevent appearance of obtained redundant features, a refinement block with signal enhancements in a serial way is designed to filter useless information. To prevent loss of original information, a multi-level enhancement mechanism guides a CNN to achieve a symmetric architecture for promoting expressive ability of HGSRCNN. Besides, a parallel up-sampling mechanism is developed to train a blind SR model. Extensive experiments illustrate that the proposed HGSRCNN has obtained excellent SR performance in terms of both quantitative and qualitative analysis. Codes can be accessed at https://***/hellloxiaotian/HGSRCNN. © 2022, CC BY-NC-SA.

关键词： Convolutional neural networks