检索结果-内蒙古大学图书馆

2020 International computer Symposium, ICS 2020

作者： Sandeep Vara Sankar, D Ko, Li-Wei National Chiao Tung University International Ph.D Program in Interdisciplinary Neuroscience Dept. of Biological Science and Technology Hsinchu Taiwan National Chiao Tung University Center for Intelligent Drug Systems and Smart Bio-Devices Department of Biological Science and Technology Hsinchu Taiwan

ISBN: (纸本)9781728192550

Fatigue is a behavioral phenomenon that occurs when a user conducts focused mental activity for a prolonged duration resulting in performance errors or lapse. This paper presents an electroencephalogram (EEG)-based fatigue and attention level detection using 2-dimensional convolutional neural networks (2D-CNN). The experimental paradigm was designed to identify and detect a target object from a multi- target scenario. The results show that our proposed model provides a classification accuracy of 86.12%, which is ~18%, ~16% and ~10% higher than the Bayesian linear discriminant analysis (BLDA), support vector machine (SVM) and bootstrap aggregating (bagging tree) algorithms. Further, the decreased session-wise accuracy levels observed for each subject after the 4th experimental session postulates cognitive state disparities were caused due to increased fatigue and dropped attention levels. These biomarkers were assessed by comparing the resting theta and alpha band powers with the rapid serial visual presentation (RSVP) performance of the later sessions (sessions 5 to 7). The results show an inverse relationship between the RSVP classification performance and the resting EEG power, validating that the subjects' performance is affected by the physiological state biomarkers like fatigue and attention for prolonged brain-computer interface (BCI) experiments. Our findings demonstrate that the resting theta and alpha band powers can be considered as indicative measures to interpret mental fatigue and attention deficit problems. © 2020 IEEE.

关键词： Brain computer interface

来源：评论

学校读者我要写书评

暂无评论

Federated Reinforcement Learning for Controlling Multiple Rotary Inverted Pendulums in Edge Computing Environments 2

Federated Reinforcement Learning for Controlling Multiple Ro...

引用

2nd International Conference on Artificial Intelligence in Information and Communication, ICAIIC 2020

作者： Lim, Hyun-Kyo Kim, Ju-Bong Kim, Chan-Myung Hwang, Gyu-Young Choi, Ho-Bin Han, Youn-Hee Korea University of Technology and Education Dept. of Interdisciplinary Program in Creative Engineering Cheon-an Korea Republic of Advanced Technology Research Center Korea University of Technology and Education Cheon-an Korea Republic of Korea University of Technology and Education Dept. of Computer Science Engineering Cheon-an Korea Republic of

ISBN: (纸本)9781728149851

Reinforcement learning has recently been studied in various fields and also used to optimally control real devices (e.g., robotic arms). In this paper, we try to allow multiple reinforcement learning agents to learn optimal control policy on their own devices of the same type but with slightly different dynamics. For such multiple devices, there is no guarantee that an agent who interacts only with one device and learns the optimal control policy will also control another device well. Therefore, we may need to apply independent reinforcement learning to each device individually, which requires time-consuming effort. To solve this problem, we propose a new federated reinforcement learning architecture where each agent working on its independent device shares their learning experience with each other, and transfers a mature policy model parameters into other agents. We incorporate the Actor-Critic PPO algorithm into each agent in the proposed collaborative architecture, and propose an efficient procedure for the gradient sharing and the model transfer. We also use edge computing to solve network problems that occur when training multiple real devices at the same time. Using multiple rotary inverted pendulum devices, we demonstrate that the proposed federated reinforcement learning scheme can effectively facilitate the learning process for multiple devices, and that the learning speed can be faster if more agents are involved. © 2020 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Quantum computers based on rare-earth compounds and PT-and anti-PT symmetric qubits

Quantum computers based on rare-earth compounds and PT-and a...

引用

Photonics for Quantum 2021

作者： Patel, Darayas N. Sarkisov, Sergey S. Darwish, Abdalla M. Mathematics and Computer Science Dept. Oakwood University 7000 Adventist Blvd HuntsvilleAL35896 United States SSS optical Technologies 515 Sparkman Drive HuntsvilleAL35816-3417 United States Physics Medical Physics and Pre-Engineering Program School of STEM Dillard University 2601 Gentilly Blvd New OrleansLA70122 United States

ISBN: (纸本)9781510645264

Hardware of modern quantum computing (QC) platforms is based mainly on superconductors and ion traps. It demands ultra-low temperature and high vacuum. Practicality and scaling up thus remain questionable. This work analyzes the alternative based on the micro- and nanoparticles of the compounds of Rare Earth (RE) elements, such as phosphors NaYF4: X3+ (X stands for Tm, Er, etc.), embedded in polymer matrices. The qubits in these systems correspond to the quantum levels of 4f electrons of RE ions, and they have optical frequencies. Qubit formation is supported by the properties of RE ions: (a) weak interaction with the environment, (b) strong inhomogeneous crystal field, and (c) the ability of neighboring ions, being in some 4f states, to interact with each other through the mechanism of Stark blockade. The latter is required for quantum conditional gate operations, such as Controlled NOT (CNOT - the basic block of a quantum computer circuit). Optical spectroscopy indicated that CNOT gate can be implemented, for instance, using the transitions in for-level system of Tm3+ ions |3H6> (ground state), |3H4> (state 0), |1D2> (state 1), and |1I6> (auxiliary state 1') activated with conventional tunable lasers at 451, 653, 798, and 1459 nm. We also considered the PT- and anti-PT-symmetry on the decoherence rate of the qubits. PT-symmetric Hamiltonian H(x) has its real part symmetric versus x-coordinate, and the anti-PT-symmetric one is anti-symmetric. Anti-PT symmetric qubits tend to be more decoherence stable than the others. It was suggested that the anti-PT-symmetry can be introduce via coupling of the qubits to anti-PT-symmetric cavities using RE-doped compounds as gain media. © 2021 Copyright SPIE.

关键词： Qubits

来源：评论

学校读者我要写书评

暂无评论

Scalable and efficient neural speech coding: A hybrid design

arXiv

引用

arXiv 2021年

作者： Zhen, Kai Sung, Jongmo Lee, Mi Suk Beack, Seungkwon Kim, Minje Department of Computer Science and Cognitive Science Program Indiana University BloomingtonIN47408 United States Electronics and Telecommunications Research Institute Daejeon34129 Korea Republic of Dept. of Intelligent Systems Engineering Indiana University United States

We present a scalable and efficient neural waveform coding system for speech compression. We formulate the speech coding problem as an autoencoding task, where a convolutional neural network (CNN) performs encoding and decoding as a neural waveform codec (NWC) during its feedforward routine. The proposed NWC also defines quantization and entropy coding as a trainable module, so the coding artifacts and bitrate control are handled during the optimization process. We achieve efficiency by introducing compact model components to NWC, such as gated residual networks and dept.wise separable convolution. Furthermore, the proposed models are with a scalable architecture, cross-module residual learning (CMRL), to cover a wide range of bitrates. To this end, we employ the residual coding concept to concatenate multiple NWC autoencoding modules, where each NWC module performs residual coding to restore any reconstruction loss that its preceding modules have created. CMRL can scale down to cover lower bitrates as well, for which it employs linear predictive coding (LPC) module as its first autoencoder. The hybrid design integrates LPC and NWC by redefining LPC’s quantization as a differentiable process, making the system training an end-to-end manner. The decoder of proposed system is with either one NWC (0.12 million parameters) in low to medium bitrate ranges (12 to 20 kbps) or two NWCs in the high bitrate (32 kbps). Although the decoding complexity is not yet as low as that of conventional speech codecs, it is significantly reduced from that of other neural speech coders, such as a WaveNet-based vocoder. For wide-band speech coding quality, our system yields comparable or superior performance to AMR-WB and Opus on TIMIT test utterances at low and medium bitrates. The proposed system can scale up to higher bitrates to achieve near transparent performance. © 2021, CC BY.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

A Computational Approach for Predicting the Termination of COVID-19

A Computational Approach for Predicting the Termination of C...

引用

Green Energy, Computing and Sustainable Technology (GECOST), International Conference on

作者： Prateek Dutta Abhiroop Sarkar Yash Ambekar Hui Ting Pek F. H. Juwono Gopal Sakarkar Department of Artificial Intelligence G H Raisoni College of Engineering Nagpur M.H India Department of Computer Science G H Raisoni College of Engineering Nagpur M.H India Dept. of Mechatronics and Biomed. Eng Universiti Tunku Abdul Rahman Kajang Selangor Malaysia Computer Science Program University of Southampton Malaysia Nusajaya Johor Malaysia D.Y. Patil Institute of Master of Computer Applications and Management Akurdi Pune M.H India

ISBN: (纸本)9781665486644

In 2019, there was an epidemic to the human society, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The virus causes coronavirus disease 2019 (COVID-19). It is an uncertain disease encountered in society for which the technology and human society had not prepared before. COVID-19 first spread over the Wuhan city of China. Since, the past two years of time-span, it has affected the citizen’s life culture and expectancy. Now, most of the population are concern about when will be COVID-19 terminate. Basically, this paper aims to analyze the COVID-19 data with features as total confirmed cases, death rate, and vaccination rate around the world-wide region. On analyzing the data, with the help of Machine Learning (ML) algorithms, we estimate the termination of COVID-19. The rapid expansion of the COVID-19 epidemic has compelled the need for technology in this field.

关键词： COVID-19 Pandemics Time series analysis Urban areas Sociology Predictive models Prediction algorithms

来源：评论

学校读者我要写书评

暂无评论

Assessing pattern recognition performance of neuronal cultures through accurate simulation

arXiv

引用

arXiv 2020年

作者： Lagani, Gabriele Mazziotti, Raffaele Falchi, Fabrizio Gennaro, Claudio Cicchini, Guido Marco Pizzorusso, Tommaso Cremisi, Federico Amato, Giuseppe Dept. of Computer Science University of Pisa Italy CNR Pisa Italy Dept. of Neuroscience University of Florence Italy Scuola Normale Superiore Pisa Italy

Previous work has shown that it is possible to train neuronal cultures on Multi-Electrode Arrays (MEAs), to recognize very simple patterns. However, this work was mainly focused to demonstrate that it is possible to induce plasticity in cultures, rather than performing a rigorous assessment of their pattern recognition performance. In this paper, we address this gap by developing a methodology that allows us to assess the performance of neuronal cultures on a learning task. Specifically, we propose a digital model of the real cultured neuronal networks;we identify biologically plausible simulation parameters that allow us to reliably reproduce the behavior of real cultures;we use the simulated culture to perform handwritten digit recognition and rigorously evaluate its performance;we also show that it is possible to find improved simulation parameters for the specific task, which can guide the creation of real cultures. © 2020, CC BY-NC-ND.

关键词： Neurons

来源：评论

学校读者我要写书评

暂无评论

An Enhanced System for Wrong-Way Driving Vehicle Detection with Road Boundary Detection Algorithm

引用

Procedia computer science 2022年 204卷 164-171页

作者： Pintusorn Suttiponpisarn Chalermpol Charnsripinyo Sasiporn Usanavasin Hiro Nakahara TAIST Tokyo Tech ICTES Program Sirindhorn International Institute of Technology Thammasat University Pathum Thani Thailand National Electronics and Computer Technology Center National Science and Technology Development Agency Pathum Thani Thailand School of ICT Sirindhorn International Institute of Technology Thammasat University Pathum Thani Thailand Dept. of Information and Communications Engineering Tokyo Institute of Technology Tokyo Japan

Driving in the wrong direction is one of the main reasons that cause road accidents in Thailand. To efficiently detect wrong direction driving vehicles, we proposed a system that can track those moving vehicles from CCTVs using deep learning and image processing techniques. Our proposed system is composed of two main algorithms: Road Lane Boundary detection from CCTV algorithm (RLB-CCTV) and Majority-Based Correct Direction Detection algorithm (MBCDD). In this paper, we introduced the RLB-CCTV algorithm and the enhanced version of MBCDD algorithms for detecting road boundaries and wrong-direction driving vehicles. Based on our experiments, our system can identify the road lane boundary and track vehicles driving in the wrong direction on each lane with an accuracy of 96.61%.

关键词： Object detection Vehicle tracking FastMOT Image processing Hough transform

来源：评论

学校读者我要写书评

暂无评论

Solvent extraction flowsheet-level modeling and simulation with radiolysis 14

Solvent extraction flowsheet-level modeling and simulation w...

引用

14th International Nuclear Fuel Cycle Conference, GLOBAL 2019 and Light Water Reactor Fuel Performance Conference, TOP FUEL 2019

作者： de Almeida, Valmor F. Azzaoui, Taha Lyon, Kevin L. University of Massachusetts Lowell Dept. of Chemical Engineering Nuclear Program LowellMA01854 United States University of Massachusetts Lowell Dept. of Computer Science and Mathematical Sciences LowellMA01854 United States Idaho National Laboratory Aqueous Separations and Radiochemistry Group Idaho FallsID83415 United States

Flowsheet-level models do not account for explicilty coupled models of water, acid, The present work is aimed at developing an extensive flowsheet simulation capability using models for solvent extraction employed in cross-cutting applications in the area of nuclear separations, safeguards, materials protection, special materials for national security and non-proliferation, critical materials, mining, and metallurgy applications. Our approach takes into account participating aqueous, organic, and vapor phases, homogeneous and heterogeneous chemical reactions, coupled with radiolysis and hydrolysis of the aqueous and organic phases. An important enabling component of the current development is the computational framework developed to allow for the dynamic coupling of multiple stages of solvent extraction. This parallel development named Cortix (https://*** [1]) facilitates the time-coupling of multiple contacting stages or multiple contacting banks of typical solvent extraction equipment. The purpose of this communication is to discuss the progress made thus far and to present preliminary results. Copyright © GLOBAL 2019 - International Nuclear Fuel Cycle Conference and TOP FUEL 2019 - Light Water Reactor Fuel Performance *** rights reserved.

关键词： Radiolysis

来源：评论

学校读者我要写书评

暂无评论

Inferring learning rules from animal decision-making 20

Inferring learning rules from animal decision-making

引用

Proceedings of the 34th International Conference on Neural Information Processing Systems

作者： Zoe C. Ashwood Nicholas A. Roy Ji Hyun Bak The International Brain Laboratory Jonathan W. Pillow Princeton Neuroscience Institute Princeton University and Dept. of Computer Science Princeton University Princeton Neuroscience Institute Princeton University Redwood Center for Theoretical Neuroscience UC Berkeley and University of California San Francisco Princeton Neuroscience Institute Princeton University and Dept. of Psychology Princeton University

ISBN: (纸本)9781713829546

How do animals learn? This remains an elusive question in neuroscience. Whereas reinforcement learning often focuses on the design of algorithms that enable artificial agents to efficiently learn new tasks, here we develop a modeling framework to directly infer the empirical learning rules that animals use to acquire new behaviors. Our method efficiently infers the trial-to-trial changes in an animal's policy, and decomposes those changes into a learning component and a noise component. Specifically, this allows us to: (i) compare different learning rules and objective functions that an animal may be using to update its policy; (ii) estimate distinct learning rates for different parameters of an animal's policy; (iii) identify variations in learning across cohorts of animals; and (iv) uncover trial-to-trial changes that are not captured by normative learning rules. After validating our framework on simulated choice data, we applied our model to data from rats and mice learning perceptual decision-making tasks. We found that certain learning rules were far more capable of explaining trial-to-trial changes in an animal's policy. Whereas the average contribution of the conventional REINFORCE learning rule to the policy update for mice learning the International Brain Laboratory's task was just 30%, we found that adding baseline parameters allowed the learning rule to explain 92% of the animals' policy updates under our model. Intriguingly, the best-fitting learning rates and baseline values indicate that an animal's policy update, at each trial, does not occur in the direction that maximizes expected reward. Understanding how an animal transitions from chance-level to high-accuracy performance when learning a new task not only provides neuroscientists with insight into their animals, but also provides concrete examples of biological learning algorithms to the machine learning community.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Self domain adapted network

arXiv

引用

arXiv 2020年

作者： He, Yufan Carass, Aaron Zuo, Lianrui Dewey, Blake E. Prince, Jerry L. Dept. of Electrical and Computer Engineering Dept. of Computer Science Johns Hopkins University BaltimoreMD21218 United States Laboratory of Behavioral Neuroscience National Institute on Aging National Institute of Health BaltimoreMD20892 United States

Domain shift is a major problem for deploying deep networks in clinical practice. Network performance drops significantly with (target) images obtained differently than its (source) training data. Due to a lack of target label data, most work has focused on unsupervised domain adaptation (UDA). Current UDA methods need both source and target data to train models which perform image translation (harmonization) or learn domain-invariant features. However, training a model for each target domain is time consuming and computationally expensive, even infeasible when target domain data are scarce or source data are unavailable due to data privacy. In this paper, we propose a novel self domain adapted network (SDA-Net) that can rapidly adapt itself to a single test subject at the testing stage, without using extra data or training a UDA model. The SDA-Net consists of three parts: adaptors, task model, and auto-encoders. The latter two are pre-trained offline on labeled source images. The task model performs tasks like synthesis, segmentation, or classification, which may suffer from the domain shift problem. At the testing stage, the adaptors are trained to transform the input test image and features to reduce the domain shift as measured by the auto-encoders, and thus perform domain adaptation. We validated our method on retinal layer segmentation from different OCT scanners and T1 to T2 synthesis with T1 from different MRI scanners and with different imaging parameters. Results show that our SDA-Net, with a single test subject and a short amount of time for self adaptation at the testing stage, can achieve significant improvements. Copyright © 2020, The Authors. All rights reserved.

关键词： Scanning

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：