检索结果-内蒙古大学图书馆

14th IEEE International Conference on Automatic Face and Gesture Recognition (FG)

作者： Zhong, Lei Bai, Chaugmin Li, Jianfeng Chen, Tong Li, Shigang Liu, Yiguang Sch Elect & Informat Engn Chongqing Key Lab Nonlinear Circuit & Intelligent Chongqing Peoples R China Hiroshima City Univ Grad Sch Informat Sci Hiroshima Japan Sichuan Univ Coll Comp Chengdu Sichuan Peoples R China

ISBN: (纸本)9781728100890

Facial expression is controlled by facial muscle and can be considered as appearance and geometric variation of key parts. One key challenging issue of static-based facial expression recognition is to capture effective information from a single facial image. In this paper, we propose a graph representation with Bidirectional RNN ( BRNN) for static-based facial expression recognition. Each node on the graph represents appearance information around the facial landmarks. Edges represent the geometric information encoded by the distance between two nodes. A bidirectional recurrent neural network utilized to process the graph extracts the appearance and geometric representation. The final representation from BRNN is fed into a fully connected layer and a Softmax layer to infer expressions. Experimental results show that this method achieves significant improvements over the state-of-art methods on three widely used facial databases ( Oulu-CASIA, CK+, and MMI), and our method reduces the error rates of the previous best methods by 42.2%, 35.9% and 18.7%, respectively.

关键词： graph structured representation facial expression recognition BRNN

来源：评论

学校读者我要写书评

暂无评论

Distributed Power Management for Dynamic Economic Dispatch in the Multimicrogrids Environment

引用

IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY 2019年第4期27卷 1651-1658页

作者： He, Xing Yu, Junzhi Huang, Tingwen Li, Chaojie Southwest Univ Coll Elect & Informat Engn Chongqing Key Lab Nonlinear Circuits & Intelligen Chongqing 400715 Peoples R China Chinese Acad Sci Inst Automat State Key Lab Management & Control Complex Syst Beijing 100190 Peoples R China Texas A&M Univ Qatar Dept Math Doha 23874 Qatar RMIT Univ Sch Elect & Comp Engn Melbourne Vic 3001 Australia

In the multimicrogrids environment, this brief proposes a distributed power management strategy to minimize a sum of generation costs function subject to generator constraints, which includes the supply-demand balance constraint, the individual constraint, the capacity constraint, and the ramp-rate constraint. First, under the framework of discontinuous dynamical system, a distributed primal-dual consensus algorithm is discussed. Then, using the property of Laplacian matrix, it is proven that the algorithm is any-time, which indicates that supply-demand balance is not disturbed. Based on theory of dynamical system, the proposed algorithm is proven to be stable, and the optimization capabilities of the distributed primal-dual consensus algorithm are also discussed. Finally, case studies based on IEEE 30-bus, IEEE-57 bus, and IEEE 300-bus system show the effectiveness and performance of the proposed distributed primal-dual consensus algorithm.

关键词： Distributed power management strategy dynamical economic dispatch (DED) multimicrogrids

来源：评论

学校读者我要写书评

暂无评论

Diffusion generalized maximum correntropy criterion algorithm for distributed estimation over multitask network

引用

DIGITAL SIGNAL PROCESSING 2018年 81卷 16-25页

作者： Chen, Feng Li, Xinyu Duan, Shukai Wang, Lidan Wu, Jiagui Southwest Univ Coll Elect & Informat Engn Chongqing 400715 Peoples R China Southwest Univ Sch Math & Stat Chongqing 400715 Peoples R China Key Lab Nonlinear Circuits & Intelligent Informat Chongqing 400715 Peoples R China

Adopting mean-square error (MSE) criterion, distributed estimation algorithms achieve desirable performance if the background noise is drawn from the Gaussian distribution. However, these algorithms may degrade considerably in non-Gaussian situations, especially for the impulsive-noise. The diffusion generalized maximum correntropy criterion (D-GMCC) algorithm is proposed in this article to address this problem. Furthermore, we contribute in multiple tasks problem, which differs from single-task estimation problem. The relatedness of tasks among nodes is also studied to uncover its impact on estimation performance. Simulations illustrate that the proposed algorithm achieves desirable performance and outperforms other related methods. The results on the mean and mean square stability analysis of the proposed algorithm are also provided. (C) 2018 Elsevier Inc. All rights reserved.

关键词： Impulsive interference Distributed estimation Generalized maximum correntropy criterion Multitask

来源：评论

学校读者我要写书评

暂无评论

Fixed-time stabilization control for port-Hamiltonian systems

引用

nonlinear DYNAMICS 2019年第2期96卷 1497-1509页

作者： Liu, Xinggui Liao, Xiaofeng Southwest Univ Coll Elect & Informat Engn Chongqing 400715 Peoples R China Yunnan Agr Univ Dept Appl Math Kunming 650201 Yunnan Peoples R China Chongqing Key Lab Nonlinear Circuits & Intelligen Chongqing 400715 Peoples R China Chongqing Univ Coll Comp Sci Chongqing 400715 Peoples R China

In this paper, the locally fixed-time and globally fixed-time stabilization problems for the port-Hamiltonian (PH) systems via the interconnection and damping assignment passivity-based control technique are discussed. The definitions of fixed-time stability region (or region of attraction) and fixed-time stability boundary are given in this paper. From this starting point, the sufficient condition of globally fixed-time attractivity of a prespecified locally fixed-time stability region is obtained. Combining the locally fixed-time stability and the globally fixed-time attractivity of a prespecified locally fixed-time stability region, the globally fixed-time stabilization problem for PH system is effectively solved. Furthermore, the globally fixed-time control scheme independent of locally fixed-time stability region has also been derived by constructing a novel Lyapunov function. A illustrative example shows that the results obtained in this paper work very well in fixed-time control design of PH systems.

关键词： Fixed-time stability region Port-Hamiltonian systems Fixed-time attractivity Stability boundary at infinity

来源：评论

学校读者我要写书评

暂无评论

Exploiting Weight-Level Sparsity in Channel Pruning with Low-Rank Approximation

Exploiting Weight-Level Sparsity in Channel Pruning with Low...

引用

IEEE International Symposium on circuits and Systems (IEEE ISCAS)

作者： Chen, Zhen Lin, Jianxin Liu, Sen Chen, Zhibo Li, Weiping Zhao, Jin Yan, Wei Univ Sci & Technol China CAS Key Lab Technol Geospatial Informat Proc & Ap Hefei Anhui Peoples R China Peking Univ Dept Integrated Circuits & Intelligent Syst Beijing Peoples R China

ISBN: (纸本)9781728103976

Acceleration and compression on Deep Neural Networks (DNNs) have become a critical problem to develop intelligence on resource-constrained hardware, especially on Internet of Things (IoT) devices. Previous works based on channel pruning can be easily deployed and accelerated without specialized hardware and software. However, weight-level sparsity is not well explored in channel pruning, which results in relatively low compression rate. In this work, we propose a framework that combines channel pruning with low-rank decomposition to tackle this problem. First, the low-rank decomposition is utilized to eliminate redundancy within filter, and achieves acceleration in shallow layers. Then, we apply channel pruning on the decomposed network in a global way, and obtains further acceleration in deep layers. In addition, a spectral norm-based indicator is proposed to balance low-rank approximation and channel pruning. We conduct a series of ablation experiments and prove that low-rank decomposition can effectively improve channel pruning by generating small and compact filters. To further demonstrate the hardware compatibility, we deploy the pruned networks on the FPGA, and the networks produced by our method have obviously low latency.

关键词： deep learning network acceleration channel pruning low-rank decomposition hardware resources

来源：评论

学校读者我要写书评

暂无评论

Diffusion least logarithmic absolute difference algorithm for distributed estimation

引用

SIGNAL PROCESSING 2018年 142卷 423-430页

作者： Chen, Feng Shi, Tao Duan, Shukai Wang, Lidan Wu, Jiagui Southwest Univ Coll Elect & Informat Engn Chongqing 400715 Peoples R China Southwest Univ Sch Math & Stat Chongqing 400715 Peoples R China Key Lab Nonlinear Circuits & Intelligent Informat Chongqing 400715 Peoples R China

The popular distributed estimation algorithms based on the mean-square error criterion is not robust against impulsive noise in the adaptive networks. To address the problem, the diffusion least logarithmic absolute difference (LLAD) algorithm is proposed in this article, which adopts both the logarithm operation and sign operation to the error. The algorithm can elegantly and gradually adjust the conventional cost functions in its optimization based on the error variation. Compared with centralized LLAD algorithm, the diffusion LLAD algorithm performs a good balance between communications and performance. The theoretical stability of mean and mean-square performance of the algorithm is analyzed. Simulation results indicate that the algorithm achieves a better performance, compared with diffusion LMS and diffusion sign-error LMS algorithms, even in the impulsive noise environment. (C) 2017 Elsevier B.V. All rights reserved.

关键词： Impulsive interference Distributed estimation Least logarithmic absolute difference

来源：评论

学校读者我要写书评

暂无评论

Energy-Efficient Dynamic Computation Offloading and Cooperative Task Scheduling in Mobile Cloud Computing

引用

IEEE TRANSACTIONS ON MOBILE COMPUTING 2019年第2期18卷 319-333页

作者： Guo, Songtao Liu, Jiadi Yang, Yuanyuan Xiao, Bin Li, Zhetao Southwest Univ Coll Elect & Informat Engn Chongqing Key Lab Nonlinear Circuits & Intelligen Chongqing 400715 Peoples R China SUNY Stony Brook Dept Elect & Comp Engn Stony Brook NY 11794 USA Hong Kong Polytech Univ Dept Comp Kowloon Hong Kong Peoples R China Xiangtan Univ Coll Informat Engn Xiangtan 411105 Hunan Peoples R China

Mobile cloud computing (MCC) as an emerging and prospective computing paradigm, can significantly enhance computation capability and save energy for smart mobile devices (SMDs) by offloading computation-intensive tasks from resource-constrained SMDs onto resource-rich cloud. However, how to achieve energy-efficient computation offloading under hard constraint for application completion time remains a challenge. To address such a challenge, in this paper, we provide an energy-efficient dynamic offloading and resource scheduling (eDors) policy to reduce energy consumption and shorten application completion time. We first formulate the eDors problem into an energy-efficiency cost (EEC) minimization problem while satisfying task-dependency requirement and completion time deadline constraint. We then propose a distributed eDors algorithm consisting of three subalgorithms of computation offloading selection, clock frequency control, and transmission power allocation. Next, we show that computation offloading selection depends on not only the computing workload of a task, but also the maximum completion time of its immediate predecessors and the clock frequency and transmission power of the mobile device. Finally, we provide experimental results in a real testbed and demonstrate that the eDors algorithm can effectively reduce EEC by optimally adjusting CPU clock frequency of SMDs in local computing, and adapting the transmission power for wireless channel conditions in cloud computing.

关键词： Mobile cloud computing energy-efficiency cost computation offloading resource allocation

来源：评论

学校读者我要写书评

暂无评论

SRMC: A Multibit Memristor Crossbar for Self-Renewing Image Mask

引用

IEEE TRANSACTIONS ON VERY LARGE SCALE INTEGRATION (VLSI) SYSTEMS 2018年第12期26卷 2830-2841页

作者： Shang, Liuting Duan, Shukai Wang, Lidan Huang, Tingwen Southwest Univ Coll Elect & Informat Engn Chongqing Key Lab Nonlinear Circuits & Intelligen Chongqing 400715 Peoples R China Southwest Univ Brain Inspired Comp & Intelligent Control Key Lab Chongqing Key Lab Nonlinear Circuits & Intelligen Coll Elect & Informat EngnChongqing Collaborat I Chongqing 400715 Peoples R China Southwest Univ Chongqing Collaborat Innovat Ctr Brain Sci Natl & Local Joint Engn Lab Intelligent Transmiss Coll Elect & Informat Engn Chongqing 400715 Peoples R China Texas A&M Univ Qatar Dept Math Doha 23874 Qatar

A recent surge of research on deep convolutional neural network (DCNN) has given a challenge on existing computation architecture which has the flows in speed and memory bottleneck. In the pretreatment of deep learning, mask operation is frequently used to remove noise or fetch information. However, under data-intensive conditions, applying mask frequently can put an extremely heavy memory/communication burden on computing system. An efficient substrate of DCNN for ameliorating mask operation is urgently needed. In this paper, we present a self-renewing mask circuit (SRMC) to alleviate/solve earlier problems. First, a new approach to apply mask is presented based on computation-in-memory architecture that implements high-performance processor and high-density memory in the same physical location. Second, we designed the peripheral circuit which can provide self-renewing function to further avoid the data exchange with an external space. As opposed to most other computational element, which calculates two 1-bit data, the proposed SRMC storage multibit value and calculate several of them parallel. The calculating ability of SRMC is significantly superior to those of computational element in general architectures. Moreover, mean filter and edge detector are implemented to illustrate the effectiveness and scalability of the proposed circuit. Finally, we discussed two possible methods to enhance the practicability of our scheme.

关键词： Computation-in-memory (CIM) image processing mask operation memristor crossbar

来源：评论

学校读者我要写书评

暂无评论

Kernel recursive maximum correntropy with Nystrom approximation

引用

NEUROCOMPUTING 2019年 329卷 424-432页

作者： Wang, Shiyuan Dang, Lujuan Qian, Guobing Jiang, Yunxiang Southwest Univ Coll Elect & Informat Engn Chongqing 400715 Peoples R China Chongqing Key Lab Nonlinear Circuits & Intelligen Chongqing 400715 Peoples R China Xi An Jiao Tong Univ Inst Artificial Intelligence & Robot Xian 710049 Shaanxi Peoples R China Shenzhen Tsinghua Univ Grad Sch Shenzhen 518055 Peoples R China

Kernel adaptive filters (KAFs) with growing network structures incur high computational burden. Generally, sparsification methods are introduced to curb the growth of the filter structure under some threshold rules, resulting in a variable structure. Unlike the sparsification methods, the Nystrom method uses a subset of data samples to form the filter structure of a fixed size. In this paper, to combat the large outliers efficiently, the kernel recursive maximum correntropy with Nystrom approximation (KRMC-NA) is proposed to achieve desirable filtering performance under a fixed and efficient filter structure. In addition, the theoretical analysis on the convergence characteristics of KRMC-NA is provided. Simulation results illustrate that the proposed KRMC-NA has better filtering accuracy than KAFs with sparsification, and approaches the filtering accuracy of the kernel recursive maximum correntropy using lower computational complexity. (C) 2018 Elsevier B.V. All rights reserved.

关键词： Kernel adaptive filters Nystrom method Kernel recursive maximum correntropy Learning

来源：评论

学校读者我要写书评

暂无评论

Fractional-order follower observer design for tracking consensus in second-order leader multi-agent systems: Periodic sampled-based event-triggered control

引用

JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS 2018年第11期355卷 4618-4628页

作者： Wang, Aijuan Liao, Xiaofeng Dong, Tao Southwest Univ Coll Elect & Informat Engn Chongqing 400715 Peoples R China Chongqing Key Lab Nonlinear Circuits & Intelligen Chongqing Peoples R China

This paper investigates the tracking consensus problem for the second-order leader systems by designing fractional-order observer, where a periodic sampled-based data event-triggered control is employed. In order to track the position information of leader, observers for followers are designed by fractional-order system, where only the relative position information is available. Furthermore, in the process of observers design, a sampled-based event-triggered strategy is proposed so that observers use the event-triggered sampled-data, to reduce the overall load of the network. In our proposed event-triggered strategy, the event detection only works at every sampling time instant which determines whether the sampled-data should be discarded or used. Under this control strategy, the Zeno-behavior is absolutely excluded since the minimum of inter-event times is inherently lower bounded by one sampling period. It is found that the followers can track state of the leader if fractional-order observers are appropriately designed and relevant parameters are properly selected. By using the generalized Nyquist stability criterion, a necessary and sufficient condition for the observer tracking consensus of the second-order leader systems is derived. The results show that the real and imaginary parts of the eigenvalues of the augmented Laplacian matrix, and fractional-order a of observer play a vital role in reaching consensus. (C) 2018 The Franklin Institute. Published by Elsevier Ltd. All rights reserved.

关键词： Multi agent systems

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：