Class Incremental ELM and Application for Image Recognition

Wei Tao,Ji Xin-Sheng 보안공학연구지원센터 2016 International Journal of Signal Processing, Image Vol.9 No.9

In image recognition field, the fact is that the trained image classifier can not recognize the images, whose class type is not the same as the training data. To resolve this problem, a new image classifier is proposed, which is based on the class incremental extreme learning machine. The new classifier can recognize the normal images well, label them with new labels, and update itself with the new labeled data. Tested on the real-world daily activity data set, the results show that our algorithm performs well.

Image Stitching with Robust Principal Component Analysis

Wei Tao,Zhang Yongxin,Yuan Yating,Ji Xinsheng 보안공학연구지원센터 2016 International Journal of Signal Processing, Image Vol.9 No.12

An image stitching algorithm based on the robustness of principal component analysis (RPCA) is proposed in an effort to suppress the influence of noise in the image stitching quality. This algorithm represents high dimensional feature data by utilizing a lower dimensional linear subspace, and converts the image stitching problem into a principal component matrix matching problem. Through the use of a low rank matrix, the extraction of salient image characteristics is recovered and the noise interference is reduced during the enhancement process. Together, with the advantages of the RPCA algorithm, the algorithm improves the PSNR of the image while maintaining its strong matching ability. Experimental results show that the proposed scheme is able to significantly inhibit the noise and improve the stitching quality in comparison to the other existing stitching methods.

Multi-phase Ge(GeO<SUB>x</SUB>)/T-Nb₂O<SUB>5-x</SUB>/C composite with synergistically improved electrochemical performance toward lithium storage

Wei Tao,Byungchan Han 한국표면공학회 2023 한국표면공학회 학술발표회 초록집 Vol.2023 No.5

Improved Quantum-Behaved Particle Swarm Method for Optimizing Complex Thin Plate Structure

Weitao Cheng,Yixiao Qin,Jinpeng Gu,Haibiao Gao,Yue Yan,Junle Yang,Yang Chen,Shen Su,Kaiyao Yang 대한토목학회 2023 KSCE Journal of Civil Engineering Vol.27 No.4

A large number of heavy-duty asymmetric thin-plate box girder structures exist in large equipment, and their optimization can reduce the amount of material used and increase their load-carrying capacity. A new optimization method based on the Improved Quantum-Behaved Particle Swarm Optimization method (IQBPSO) is proposed in order to efficiently solve the mathematical model for the rationalization and optimization design of structures. The penalty function and Lévy flight strategy are considered in the optimization design of the improved algorithm, thus transforming the constrained optimization problem into an unconstrained optimization problem and improving the diversity and local optimization search capability of the quantum particle swarm. A mathematical model for the optimal design of box girder section size is established with the reduction of beam cross-sectional area as the objective function and the thin plate strength, rigidity, and stability of the thin slab as the constraints. The rapid lightweight design of the thin plate box beam was achieved, resulting in a 9.6% reduction in the manufacturing cost of the thin plate box beam. The optimization results are compared with several solutions of the thin slab box beams to verify the reliability and validity of the proposed optimization method.

Reliability-based design optimization using reliability mapping functions

Weitao Zhao,Xueyan Shi,Kai Tang 국제구조공학회 2017 Structural Engineering and Mechanics, An Int'l Jou Vol.62 No.2

Reliability-based design optimization (RBDO) is a powerful tool for design optimization when considering probabilistic characteristics of design variables. However, it is often computationally intensive because of the coupling of reliability analysis and cost minimization. In this study, the concept of reliability mapping function is defined based on the relationship between the reliability index obtained by using the mean value first order reliability method and the failure probability obtained by using an improved response surface method. Double-loop involved in the classical RBDO can be converted into single-loop by using the reliability mapping function. Since the computational effort of the mean value first order reliability method is minimal, RBDO by using reliability mapping functions should be highly efficient. Engineering examples are given to demonstrate the efficiency and accuracy of the proposed method. Numerical results indicated that the proposed method has the similar accuracy as Monte Carlo simulation, and it can obviously reduce the computational effort.

An efficient response surface method considering the nonlinear trend of the actual limit state

Weitao Zhao,Zhiping Qiu,Yi Yang 국제구조공학회 2013 Structural Engineering and Mechanics, An Int'l Jou Vol.47 No.1

In structural reliability analysis, the response surface method is a powerful method to evaluate the probability of failure. However, the location of experimental points used to form a response surface function must be selected in a judicious way. It is necessary for the highly nonlinear limit state functions to consider the design point and the nonlinear trend of the limit state, because both of them influence the probability of failure. In this paper, in order to approximate the actual limit state more accurately, experimental points are selected close to the design point and the actual limit state, and consider the nonlinear trend of the limit state. Linear, quadratic and cubic polynomials without mixed terms are utilized to approximate the actual limit state. The direct Monte Carlo simulation on the approximated limit state is carried out to determine the probability of failure. Four examples are given to demonstrate the efficiency and the accuracy of the proposed method for both numerical and implicit limit states.

Fault injection and failure analysis on Xilinx 16 nm FinFET Ultrascale+ MPSoC

Weitao Yang,Yonghong Li,Chaohui He 한국원자력학회 2022 Nuclear Engineering and Technology Vol.54 No.6

Energetic particle strikes the device and induces data corruption in the configuration memory (CRAM),causing errors and even malfunctions in a system on chip (SoC). Software-based fault injection is aconvenient way to assess device performance. In this paper, dynamic partial reconfiguration (DPR) isadopted to make fault injection on a Xilinx 16 nm FinFET Ultrascaleþ MPSoC. And the reconfigurationmodule implements the Sobel and Gaussian image filtering, respectively. Fault injections are executed onthe static and reconfiguration modules' bitstreams, respectively. Another contribution is that the failuremodes and effects analysis (FMEA) method is applied to evaluate the system reliability, according to theobtained injection results. This paper proposes a software-based solution to estimate programmabledevice vulnerability.

An Improvement of the Response Surface Method Based on Reference Points for Structural Reliability Analysis

Weitao Zhao,Weihua Liu,Qijiao Yang 대한토목학회 2016 KSCE Journal of Civil Engineering Vol.20 No.7

The Response Surface Method (RSM) is a powerful technique to evaluate the structural reliability. However, for a Limit State Function (LSF) with highly non-linear, the accuracy of the approximation of the failure probability does not depend very much upon the design point. It is necessary for RSMs to consider the design point and the non-linear trend of actual LSF around the design point, because both of them influence the failure probability. Thus, in order to improve the fitting precision of the Response Surface Function (RSF) to the actual LSF over a larger region containing the design point, the reference points of experimental points are constructed in this paper. Experimental points used to obtain parameters of a RSF are selected according to the information of reference points. Four examples are discussed in detail. The numerical results indicate that the accuracy and the efficiency of the proposed method are both desirable for both numerical and implicit LSFs, and the proposed method is superior to the classical RSM in terms of efficiency and accuracy.

Comparison of changes in the nasal cavity, pharyngeal airway, and maxillary sinus volumes after expansion and maxillary protraction with two protocols: Rapid palatal expansion versus alternate rapid maxillary expansion and constriction

Weitao Liu,Shaonan Zhou,Edwin Yen,Bingshuang Zou 대한치과교정학회 2023 대한치과교정학회지 Vol.53 No.3

Objective: To evaluate and compare a series of volume changes in the nasal cavity (NC), nasopharynx, oropharynx, and maxillary sinuses (MS) in growing Class III patients after either rapid palatal expansion (RPE) or alternate rapid maxillary expansion and constriction (Alt-RAMEC) followed by facemask (FM) therapy, by using cone-beam computed tomography (CBCT). Methods: Forty growing Class III patients were retrospectively selected and divided into two matched groups: RPE/FM (14 females, 6 males; mean age, 9.66 ± 1.23 years) and Alt-RAMEC/FM groups (14 females, 6 males; mean age, 10.28 ± 1.45 years). The anteroposterior and vertical displacements of Point A, the volumes of the NC, nasopharyngeal, oropharyngeal, and MS were measured at different time points: pretreatment (T1), postexpansion (T2), and postprotraction (T3). Results: Both groups demonstrated significant maxilla advancement (by 1.3 mm) during expansion, with a statistically significant intergroup difference during protraction (RPE/FM, 1.1 mm; Alt-RAMEC/FM, 2.4 mm; p < 0.05) and throughout the treatment (RPE/FM, 2.4 mm; Alt-RAMEC/FM, 3.7 mm; p < 0.05). NC and nasopharyngeal airway volumes increased significantly in both groups after expansion, protraction, and treatment. The oropharyngeal and MS volumes increased in both groups after protraction and post-treatment. However, no volumetric differences were observed between the two groups. Conclusions: There was no significant difference in airway volume changes, including NC, nasopharyngeal, oropharyngeal airway, and MS, between RPE/FM and Alt-RAMEC/FM groups at different time points. Although there was significantly more forward movement after protraction in the Alt-RAMEC/FM group, the difference was deemed too small to be clinically relevant.

Food intake and its effect on the species and abundance of intestinal flora in colorectal cancer and healthy individuals

( Weitao Shen ),( Jiayu Sun ),( Zhiyang Li ),( Fen Yao ),( Kaihuang Lin ),( Xiaoyang Jiao ) 대한내과학회 2021 The Korean Journal of Internal Medicine Vol.36 No.3

Background/Aim: It is known that an imbalance in the intestinal f lora plays a crucial role in colorectal cancer (CRC), but the effect of food consumption patterns on the types of intestinal flora remains to be clarified. We aimed to analyze the associations between food intake and intestinal flora in healthy and CRC individuals. Methods: Food intake data were recorded using the Food Frequency Questionnaire (FFQ). The composition and diversity of the intestinal flora detected by 16S rRNA gene sequencing, and the data were analyzed by R version 3.1.1 software. Results: Higher intake of red meat or pickled foods, and lower intake of white meat, fruits, vegetables, beans, nuts were found in the CRC group compared with the healthy group. Higher levels of Fusobacteria and Proteobacteria, and lower levels of Firmicutes were observed in the CRC group. Partial correlation analysis revealed that the intake of fruits, beans, and nuts was negatively correlated with Proteobacteria and Fusobacteria, but pickled food was positively correlated with Fusobacteria (p < 0.05). Fish, beans, and nuts intake was negatively correlated with Escherichia (p = 0.01). Multiple regression analysis revealed that vegetable oil (odds ratio [OR], 0.26; 95% confidence interval [CI], 0.13 to 0.82), vegetables (OR, 0.26; 95% CI, 0.10 to 0.64), eggs (OR, 0.26; 95% CI, 0.10 to 0.69), pickled foods (OR, 21.02; 95% CI, 6.02 to 73.45), and red meat (OR, 4.23; 95% CI, 1.68 to 10.60) had an impact on CRC risk. Conclusions: The species and abundance of intestinal flora varies between CRC and healthy individuals and may be affected by their food preference.