Yi Xu

Yi Xu


Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 February 2019 in Sensors
Verified
Journal Article
Open access
PDF file
YXYi Xu
Yi Xu
GSGuiling Sun
Guiling Sun
TGTIANYU GENG
TIANYU GENG
Jingfei HeJingfei He
Jingfei He
Sparse sensing schemes based on matrix completion for data collection have been proposed to reduce the power consumption of data-sensing and transmission in wireless sensor networks (WSNs). While extensive efforts have been made to improve the recovery accuracy from the sparse samples, it is usually at the cost of running time. Moreover, most data-collection methods are difficult to implement with low sampling ratio because of the communication limit. In this paper, we design a novel data-collection method including a Rotating Random Sparse Sampling method and a Fast Singular Value Thresholding algorithm. With the proposed method, nodes are in the sleep mode most of the time, and the sampling ratio varies over time slots during the sampling process. From the samples, a corresponding algorithm with Nesterov technique is given to recover the original data accurately and fast. With two real-world data sets in WSNs, simulations verify that our scheme outperforms other schemes in terms of energy consumption, reconstruction accuracy, and rate. Moreover, the proposed sampling method enhances the recovery algorithm and prolongs the lifetime of WSNs.

23 October 2019 in IEEE Access
Verified
Journal Article
Open access
YXYi Xu
Yi Xu
GSGuiling Sun
Guiling Sun
TGTianyu Geng
Tianyu Geng
BZBowen Zheng
Bowen Zheng
Wireless Sensor Networks (WSNs) have been deeply studied by many researchers and been widely used in many fields. Since a large amount of energy for WSNs is used for sensing and transmitting, researchers come up with many methods to reduce the number of sensed and transmitted data packets. Compressive Data Gathering (CDG) is a well-known method to gather WSNs data, but it does not realize sparse sensing as it needs to sense all data and compress them. The efficiency of Low-rank and TV regularizations for recovering WSNs data has been demonstrated, however, they are not combined to enable utilization of data correlation throughout the network. To recover the data accurately and to reduce the energy consumption in WSNs, we propose a Compressive Sparse Data Gathering (CSDG) scheme including a Compressive Sparse Sampling (CSS) method and a data recovery algorithm based on low-rank and Total Variation (TV) regularizations fully exploiting the sparsity and low-rank characteristics of WSNs data. The alternating direction method of multipliers and the steepest gradient descent method are used to solve the problem. Simulations show that the CSDG method outperforms the state-of-the-art methods in terms of the recovery accuracy. Moreover, with fairly low sparse sampling ratio and high compression ratio, CSDG method can still recover the original signal with little error. As the number of sensed data and transmitted data is reduced greatly with sparse sampling and compression, the energy consumption of WSNs is lessen and the lifetime is prolonged.

13 September 2019 in IEEE Communications Letters
Verified
Journal Article
Closed access
Jingfei HeJingfei He
Jingfei He
YZYatong Zhou
Yatong Zhou
GSGuiling Sun
Guiling Sun
YXYi Xu
Yi Xu
For heterogeneous WSNs with various types of sensors, compressive data gathering method requires more measurements due to the increased multiple attributes. In this letter, a compressive multi-attribute data gathering method using low-rank Hankel matrix is proposed to reduce the required measurements and improve the recovery accuracy in heterogeneous WSNs. Beyond utilizing just the spatiotemporal correlation of the raw sensed data with compressed sensing, the proposed method further enforces the low-rank block Hankel matrix to exploit the inherent correlation among multi-attribute data. Experimental results demonstrate that the proposed method can significantly improve the recovery accuracy of multi-attribute data compared with the existing solutions in WSNs.

13 July 2020 in arXiv
Verified
Preprint
Open access
PDF file
YYYan Yan
Yan Yan
YXYi Xu
Yi Xu
LZLijun Zhang
Lijun Zhang
XWXiaoyu Wang
Xiaoyu Wang
TYTianbao Yang
Tianbao Yang
In this paper, we study a family of non-convex and possibly non-smooth inf-projection minimization problems, where the target objective function is equal to minimization of a joint function over another variable. This problem include difference of convex (DC) functions and a family of bi-convex functions as special cases. We develop stochastic algorithms and establish their first-order convergence for finding a (nearly) stationary solution of the target non-convex function under different conditions of the component functions. To the best of our knowledge, this is the first work that comprehensively studies stochastic optimization of non-convex inf-projection minimization problems with provable convergence guarantee. Our algorithms enable efficient stochastic optimization of a family of non-decomposable DC functions and a family of bi-convex functions. To demonstrate the power of the proposed algorithms we consider an important application in variance-based regularization. Experiments verify the effectiveness of our inf-projection based formulation and the proposed stochastic algorithm in comparison with previous stochastic algorithms based on the min-max formulation for achieving the same effect.

28 March 2023 in Teacher Education and Curriculum Studies
Verified
Journal Article
Open access
PDF file
YXYi Xu
Yi Xu
NAO is the most well-known and widely used humanoid robot in education and research in the world. With the development of computer technology and cloud computing, and the establishment of artificial intelligence specialty, robot teaching is more important in artificial intelligence specialty. The artificial intelligence specialty is a new specialty just established in the talent training program of Chinese universities. Its purpose is to cultivate application-oriented talents in China's artificial intelligence industry and promote the construction of first-class disciplines of artificial intelligence. MATLAB is the commercial mathematical software. It is a high-level technical computing language and interactive environment for algorithm development, data visualization, data analysis and numerical calculation. It is widely used in the experimental teaching of NAO robot. In the teaching of robot theory, kinematics, trajectory planning, dynamics and control are relatively abstract and complex, and there are a lot of formula derivation and differential equations. In order to improve the quality of NAO experiment teaching, this paper analyzes the current situation and existing problems in NAO experiment teaching. In the exploration and practice of robot experiment teaching, NAO robot is taken as the research object, and MATLAB is used to carry out robot simulation and control experiments. The paper presents a multi-level experimental teaching method of NAO robot based on the combination of virtual and Reality. The multi-level includes three levels: basic experiment, design experiment and innovative experiment. The paper describes the design and implementation process of the three-level experimental teaching in detail, and gives the assessment methods, characteristics and effectiveness evaluation of the experimental teaching. The teaching results show that the teaching effect is good, can stimulate students' enthusiasm for learning, and help to cultivate students' programming ability and innovation ability.