检索结果-内蒙古大学图书馆

Low-latency federated learning over wireless channels with differential privacy

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Wei, Kang Li, Jun Ma, Chuan Ding, Ming Chen, Cailian Jin, Shi Han, Zhu Poor, H. Vincent School of Electrical and Optical Engineering Nanjing University of Science and Technology Nanjing210094 China Data61 CSIRO SydneyNSW2015 Australia The Department of Automation Shanghai Jiao Tong University Shanghai200240 China Key Laboratory of System Control and Information Processing Ministry of Education of China Shanghai200240 China The National Mobile Communications Research Laboratory Southeast University Nanjing210096 China The Department of Electrical and Computer Engineering The University of Houston HoustonTX77004 United States The Department of Computer Science and Engineering Kyung Hee University Seoul Korea Republic of The Department of Electrical and Computer Engineering Princeton University PrincetonNJ08544 United States

In federated learning (FL), model training is distributed over clients and local models are aggregated by a central server. The performance of uploaded models in such situations can vary widely due to imbalanced data distributions, potential demands on privacy protections, and quality of transmissions. In this paper, we aim to minimize FL training delay over wireless channels, constrained by overall training performance as well as each client's differential privacy (DP) requirement. We solve this problem in the framework of multi-agent multiarmed bandit (MAMAB) to deal with the situation where there are multiple clients confornting different unknown transmission environments, e.g., channel fading and interferences. Specifically, we first transform the long-term constraints on both training performance and each client's DP into a virtual queue based on the Lyapunov drift technique. Then, we convert the MAMAB to a max-min bipartite matching problem at each communication round, by estimating rewards with the upper confidence bound (UCB) approach. More importantly, we propose two efficient solutions to this matching problem, i.e., modified Hungarian algorithm and greedy matching with a better alternative (GMBA), in which the first one can achieve the optimal solution with a high complexity while the second one approaches a better tradeoff by enabling a verified low-complexity with little performance loss. In addition, we develop an upper bound on the expected regret of this MAMAB based FL framework, which shows a linear growth over the logarithm of communication rounds, justifying its theoretical feasibility. Extensive experimental results are conducted to validate the effectiveness of our proposed algorithms, and the impacts of various parameters on the FL performance over wireless edge networks are also discussed. Copyright © 2021, The Authors. All rights reserved.

关键词： Multi agent systems

Pro-UIGAN: Progressive Face Hallucination from Occluded Thumbnails

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Zhang, Yang Yu, Xin Lu, Xiaobo Liu, Ping The Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information Ministry of Education School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China The Australian Institute of Artificial Intelligence University of Technology Sydney UltimoNSW2007 Australia The School of Automation Southeast University Nanjing210096 China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing210096 China Singapore138634 Singapore

In this paper, we study the task of hallucinating an authentic high-resolution (HR) face from an occluded thumbnail. We propose a multi-stage Progressive Upsampling and Inpainting Generative Adversarial Network, dubbed Pro-UIGAN, which exploits facial geometry priors to replenish and upsample (8×) the occluded and tiny faces (16×16 pixels). Pro-UIGAN iteratively (1) estimates facial geometry priors for low-resolution (LR) faces and (2) acquires non-occluded HR face images under the guidance of the estimated priors. Our multi-stage hallucination network upsamples and inpaints occluded LR faces via a coarse-to-fine fashion, significantly reducing undesirable artifacts and blurriness. Specifically, we design a novel cross-modal attention module for facial priors estimation, in which an input face and its landmark features are formulated as queries and keys, respectively. Such a design encourages joint feature learning across the input facial and landmark features, and deep feature correspondences will be discovered by attention. Thus, facial appearance features and facial geometry priors are learned in a mutually beneficial manner. Extensive experiments show that our Pro-UIGAN attains visually pleasing completed HR faces, thus facilitating downstream tasks, i.e., face alignment, face parsing, face recognition as well as expression classification. Copyright © 2021, The Authors. All rights reserved.

关键词： Generative adversarial networks

Integral control barrier function for dynamically defined control laws

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Ames, Aaron D. Notomista, Gennaro Wardi, Yorai Egerstedt, Magnus Mechanical and Civil Engineering Control and Dynamical Systems Caltech PasadenaCA91125 United States George W. Woodruff School of Mechanical Engineering Georgia Tech Atlanta United States School of Electrical and Computer Engineering Georgia Tech AtlantaGA30332 United States

This paper introduces integral control barrier functions (I-CBFs) as a means to enable the safety-critical integral control of nonlinear systems. Importantly, I-CBFs allow for the holistic encoding of both state constraints and input bounds in a single framework. We demonstrate this by applying them to a dynamically defined tracking controller, thereby enforcing safety in state and input through a minimally invasive I-CBF controller framed as a quadratic program. Copyright © 2020, The Authors. All rights reserved.

关键词： Quadratic programming

Attribute rule performance in data mining for software deformity prophecy datasets models

学校读者我要写书评

暂无评论

Attribute rule performance in data mining for software defor...

2019 International Conference on Advances in the Emerging Computing Technologies, AECT 2019

作者： Shaikh, Salahuddin Changan, Liu Malik, Maaz Rasheed North China Electric Power Universty School of Control Computer Engineeringt Beijing China Guilin University of Electronic Technology Dept. of Information Communication Engineering Guilin China

ISBN: (纸本)9781728144528

In recently, all the developers, programmer and software engineers, they are working specially on software component and software testing to compete the software technology in the world. For this competition, they are using different kind of sources to analysis the software reliability and importance. Nowadays Data mining is one of source, which is used in software for overcome the problem of software fault which occur during the software test and its analysis. This kind of problem leads software deformity prophecy in software. In this research paper, we are also trying to overcome the software deformity prophecy problem with the help of our proposed solution called ONER rule attribute. We have used REPOSITORY datasets models, these datasets models are defected and non-defected datasets models. Our analysis class of interest is defected models. In our research, we have analyzed the efficiency of our proposed solution methods. The experiments results showed that using of ONER with discretize, have improved the efficiency of correctly classified instances in all. Using percentage split and training datasets with ONER discretize rule attribute have improved correctly classified in all datasets models. The analysis of positive accuracy f-measure is also increased in percentage split during the use of ONER with discretize but in some datasets models, the training data and cross validation is better with use of ONER rule attribute. The area under curve (ROC) in both scenarios using ONER rule attribute and discretize with ONER rule attribute is almost same or equal with each other. © 2020 IEEE.

关键词： Data mining

Graph attention layer evolves semantic segmentation for road pothole detection: A benchmark and algorithms

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Fan, Rui Wang, Hengli Wang, Yuan Liu, Ming Pitas, Ioannis Department of Control Science and Engineering College of Electronics and Information Engineering Tongji University Shanghai201804 China Shanghai Research Institute for Intelligent Autonomous Systems Shanghai201210 China Department of Electronic and Computer Engineering The Hong Kong University of Science and Technology Hong Kong SAR Hong Kong Industrial R&d Center SmartMore Shenzhen518000 China School of Informatics University of Thessaloniki Thessaloniki541 24 Greece

Existing road pothole detection approaches can be classified as computer vision-based or machine learning-based. The former approaches typically employ 2-D image analysis/ understanding or 3-D point cloud modeling and segmentation algorithms to detect (i.e., recognize and localize) road potholes from vision sensor data, e.g., RGB images and/or depth/disparity images. The latter approaches generally address road pothole detection using convolutional neural networks (CNNs) in an end-to-end manner. However, road potholes are not necessarily ubiquitous and it is challenging to prepare a large wellannotated dataset for CNN training. In this regard, while computer vision-based methods were the mainstream research trend in the past decade, machine learning-based methods were merely discussed. Recently, we published the first stereo visionbased road pothole detection dataset and a novel disparity transformation algorithm, whereby the damaged and undamaged road areas can be highly distinguished. However, there are no benchmarks currently available for state-of-the-art (SoTA) CNNs trained using either disparity images or transformed disparity images. Therefore, in this paper, we first discuss the SoTA CNNs designed for semantic segmentation and evaluate their performance for road pothole detection with extensive experiments. Additionally, inspired by graph neural network (GNN), we propose a novel CNN layer, referred to as graph attention layer (GAL), which can be easily deployed in any existing CNN to optimize image feature representations for semantic segmentation. Our experiments compare GAL-DeepLabv3+, our best-performing implementation, with nine SoTA CNNs on three modalities of training data: RGB images, disparity images, and transformed disparity images. The experimental results suggest that our proposed GAL-DeepLabv3+ achieves the best overall pothole detection accuracy on all training data modalities. The source code, dataset, and benchmark are publicly available at m

关键词： Landforms

A biologically inspired feature enhancement framework for zero-shot learning

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Xie, Zhongwu Cao, Weipeng Wang, Xizhao Ming, Zhong Zhang, Jingjing Zhang, Jiyong College of Computer Science and Software Engineering Shenzhen University Shenzhen China College of Command and Control Engineering Army Engineering University of People’s Liberation Army of China Nanjing China School of Automation Hangzhou Dianzi University Hangzhou China

Most of the Zero-Shot Learning (ZSL) algorithms currently use pre-trained models as their feature extractors, which are usually trained on the ImageNet data set by using deep neural networks. The richness of the feature information embedded in the pre-trained models can help the ZSL model extract more useful features from its limited training samples. However, sometimes the difference between the training data set of the current ZSL task and the ImageNet data set is too large, which may lead to the use of pre-trained models has no obvious help or even negative impact on the performance of the ZSL model. To solve this problem, this paper proposes a biologically inspired feature enhancement framework for ZSL. Specifically, we design a dual-channel learning framework that uses auxiliary data sets to enhance the feature extractor of the ZSL model and propose a novel method to guide the selection of the auxiliary data sets based on the knowledge of biological taxonomy. Extensive experimental results show that our proposed method can effectively improve the generalization ability of the ZSL model and achieve state-of-the-art results on three benchmark ZSL tasks. We also explained the experimental phenomena through the way of feature visualization. Copyright © 2020, The Authors. All rights reserved.

关键词： Taxonomies

A Biologically Inspired Feature Enhancement Framework for Zero-Shot Learning

学校读者我要写书评

暂无评论

A Biologically Inspired Feature Enhancement Framework for Ze...

IEEE International Conference on Cyber Security and Cloud Computing (CSCloud)

作者： Zhongwu Xie Weipeng Cao Xizhao Wang Zhong Ming Jingjing Zhang Jiyong Zhang College of Computer Science and Software Engineering Shenzhen University Shenzhen China College of Command and Control Engineering Army Engineering University of People’s Liberation Army of China Nanjing China School of Automation Hangzhou Dianzi University Hangzhou China

ISBN: (数字)9781728165509

ISBN: (纸本)9781728165516

Most of the Zero-Shot Learning (ZSL) algorithms currently use pre-trained models as their feature extractors, which are usually trained on ImageNet by using deep neural networks. The richness of the feature information embedded in the pre-trained models can help the ZSL model extract more useful features from its limited training data. However, sometimes the difference between the training data of the current ZSL task and the ImageNet is too large, which may cause the use of pre-trained models has no obvious help or even negative impact on the model performance. To solve this problem, this paper proposes a biologically inspired feature enhancement framework for ZSL. Specifically, we design a dual-channel learning framework that uses auxiliary data sets to enhance the feature extractor of the ZSL model and propose a novel method to guide the selection of the auxiliary data sets based on the knowledge of biological taxonomy. Extensive experimental results show that our proposed method can effectively improve the generalization ability of the ZSL model and achieve state-of-the-art results on three benchmark ZSL tasks. We also explained the experimental phenomena through the way of feature visualization.

关键词： Zero-shot learning feature enhancement feature transfer biological taxonomy ancillary data Respiratory Inspiration extraction device Learning Frameworks Training data Refractive surgery enhancement

Learning Greenhouse Climate control Policy from Monitored Data

学校读者我要写书评

暂无评论

Learning Greenhouse Climate Control Policy from Monitored Da...

Chinese Automation Congress (CAC)

作者： Xiaoxuan Zhao Haoyu Wang Xiujuan Wang Udom Lewlomphaisarl Dong Li Jing Hua Mengzhen Kang The State Key Laboratory for Management and Control of Complex Systems Chinese Academy of Sciences Institute of Automation Beijing China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing China Beijing Engineering Research Center of Intelligent Systems and Technologies Beijing China National Electronics and Computer Technology Center Pathum Thani Thailand Institute of Digital Agriculture Zhejiang Academy of Agricultural Sciences Hangzhou China Qingdao AgriTech Co. Ltd. Qingdao China

The knowledge of solar greenhouse growers on environment control plays an important role in greenhouse production and management. We proposed a way to extract the control strategies from the monitored data of greenhouses by building a long short-term memory (LSTM) model. The result is verified according to the real monitored data of a solar greenhouse, which shows that the model can learn the control strategy of a ventilator in the solar greenhouse. Through monitored data and models, the knowledge of greenhouse ventilation control can be learned, and automatic control can be achieved in a greenhouse with a similar configuration.

关键词： Ventilators Green buildings Buildings Production Ventilation Data models Data mining

Electrical and Mechanical Sensor-Based Mass Flow Rate Measurement System: A Comparative Approach

学校读者我要写书评

暂无评论

Electrical and Mechanical Sensor-Based Mass Flow Rate Measur...

International Conference on Robotics and Automation Sciences (ICRAS)

作者： Usama Abrar Liu Shi Nasif Raza Jaffri Michael Short Khawar Hasham School of Control and Computer Engineering North China Electric Power University Beijing 102206 China School of Computing Engineering and Digital Technologies Teesside University Middlesbrough Cleveland TS1 3BA UK Interactive Group Islamabad Pakistan

ISBN: (数字)9781728198576

ISBN: (纸本)9781728198583

Gas-Solid flow finds its application in various industries such as chemical and food processes, pharmaceuticals, automobile and power generation. The precise measurement and real-time monitoring of the flow process have garnered a lot of attraction in the modern industrial era. The effective utilization, with high efficiency, of the resources have compelled the researchers to investigate Gas-Solid flow monitoring and measurement process for the last two decades. Efforts are extended to design instruments for flow pattern visualization and quantification of mass flow rate. Generally, volumetric flow concentration and velocity profile of the solid particles are required to find the mass flow rate (MRF) in the case of non-invasive measurement techniques. This paper proposes a novel non-invasive mass flow rate measurement system for solid particles where electrical and mechanical sensors are simultaneously used to compensate each other's associated drawbacks. The solid particles are allowed to slide along an insulated inclined channel under the effect of gravity. The velocity profile and volumetric flow concentration are calculated using electrical sensors. Furthermore, load cell serves as a second sensor to calculate the mass flow rate. The experimental results from both the sensors are compared to investigate the accuracy and relative errors of both the sensors under different conditions.

关键词：