http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Binary Hashing CNN Features for Action Recognition
( Weisheng Li ),( Chen Feng ),( Bin Xiao ),( Yanquan Chen ) 한국인터넷정보학회 2018 KSII Transactions on Internet and Information Syst Vol.12 No.9
The purpose of this work is to solve the problem of representing an entire video using Convolutional Neural Network (CNN) features for human action recognition. Recently, due to insufficient GPU memory, it has been difficult to take the whole video as the input of the CNN for end-to-end learning. A typical method is to use sampled video frames as inputs and corresponding labels as supervision. One major issue of this popular approach is that the local samples may not contain the information indicated by the global labels and sufficient motion information. To address this issue, we propose a binary hashing method to enhance the local feature extractors. First, we extract the local features and aggregate them into global features using maximum/minimum pooling. Second, we use the binary hashing method to capture the motion features. Finally, we concatenate the hashing features with global features using different normalization methods to train the classifier. Experimental results on the JHMDB and MPII-Cooking datasets show that, for these new local features, binary hashing mapping on the sparsely sampled features led to significant performance improvements.
Sampled-data Based Distributed Convex Optimization with Eventtriggered Communication
Weisheng Chen,Jiayun Liu,Hao Dai 제어·로봇·시스템학회 2016 International Journal of Control, Automation, and Vol.14 No.6
This paper studies the distributed convex optimization problem for multi-agent systems over undirectedand connected networks. Motivated by practical considerations, we propose a new distributed optimization algorithmwith event-triggered communication. The proposed event detection is decentralized, sampled-data and notrequires periodic communications among agents to calculate the threshold. Based on Lyapunov approaches, weshow that the proposed algorithm is asymptotically converge to the unknown optimizer if the design parameters arechosen properly. We also give an upper bound on the convergence rate. Finally, we illustrate the effectiveness ofthe proposed algorithm by a numerical simulation.
Weisheng Chen,Li Zhang 제어·로봇·시스템학회 2010 International Journal of Control, Automation, and Vol.8 No.2
First of all, an adaptive iterative learning control strategy is developed for a class of nonlinearly parameterized systems with two unknown time-varying parameters and one unknown time-varying delay. The proposed control law includes a PID-type feedback term in time domain and an adaptive learning term used to estimate the unknown time-varying vector in iteration domain. By constructing a Lyapunov-Krasovskii-like composite energy function, we prove the stability of the closed-loop system and the convergence of the tracking error. Then, the design idea is further extended to a broader class of systems with mixed parameters in which the unknown time-invariant vector is estimated by a PI-type learning law in time domain. The simulation results, for a time-delay chaotic system, confirm the effectiveness of the proposed control scheme.
Moving long baseline positioning algorithm with uncertain sound speed
Weisheng Yan,Wei Chen,Rongxin Cui 대한기계학회 2015 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.29 No.9
This paper presents a Moving long baseline (MLBL) positioning algorithm for the underwater target considering the uncertainunderwater sound speed. First, the positioning and the sound speed models are established. To tackle the uncertain sound speed, auncertain Least squares (ULS) positioning algorithm is applied to estimate the target position and the sound speed. Then it is essentiallyshown that four mobile buoys are necessarily (at least) required to locate the target. Further, it is found that under a singularityscenario in which the ranges between the target and each of the mobile buoys are equal, there is no solution to the positioning accordingto the classical geometrical equations. In order to solve this singularity problem, an ULS-based unscented Kalman filter(UKF) algorithm is proposed to obtain the estimated solution. Simulation results illustrate the effectiveness of proposed methods.
Jian Wu,Weisheng Chen,Jingli Du,Wenlong Ren 제어·로봇·시스템학회 2013 International Journal of Control, Automation, and Vol.11 No.5
This paper addresses the stabilization problem for a class of stochastic nonlinear systems with arbitrary switching. Based on the simultaneous domination approach, the common Lyapunov function method and the backstepping technique, a state feedback controller and an output feedback controller are designed, respectively. The closed-loop systems are proved to be globally asymptotically stable in probability. The main advantage of the proposed control schemes is that the controllers are independent of switching signal. Two simulation examples are given to illustrate the effectiveness of the proposed control strategies.
Detecting small lung tumors in mouse models by refractive-index microradiology
Chien, Chia-Chi,Zhang, Guilin,Hwu, Y.,Liu, Ping,Yue, Weisheng,Sun, Jianqi,Li, Yan,Xue, Hongjie,Xu, Lisa X.,Wang, Chang Hai,Chen, Nanyow,Lu, Chien Hung,Lee, Ting-Kuo,Yang, Yuh-Cheng,Lu, Yen-Ta,Ching, Y Springer-Verlag 2011 ANALYTICAL AND BIOANALYTICAL CHEMISTRY Vol.401 No.3
Influence of different silica gelling agents on the performance of aqueous gel electrolytes
Changyu Lu,Tuan K.A. Hoang,The Nam Long Doan,The Nam Long Doan,Matthew Acton,WEISHENG GUAN,P. Chen 한국공업화학회 2016 Journal of Industrial and Engineering Chemistry Vol.42 No.-
In this study, gel electrolytes are prepared from silica gelling agents and aqueous electrolyte forapplication in the rechargeable hybrid aqueous battery for the first time. The required quantities of silicamaterials are in the range of 4–15 wt.%. The gelling time is in the magnitude of minutes and the ionicconductivities of gel electrolytes are in the range of 50.6–60.6 mS cm 1. These batteries using these gelelectrolytes exhibit higher rate capability, slower self-discharge (ca.15% after 24 h), and up to 10–12%higher cyclability compared with the performance of the batteries using the conventional liquidelectrolyte.