matlab创新奖代码-CuiZewen1.github.io:ZewenCui

 # ZEWEN CUI (Kevin) [CV](http://github.com/cuizewen/cuizewen.github.io/raw/master/CV.pdf) [E-mail](cuizewen2019@outlook.com) [Transcript](http://github.com/cuizewen/cuizewen.github.io/raw/master/2333.jpg) ### EDUCATION **University of Electronic Science and Technology of China(UESTC)** B.Eng., Electronic science and technology GPA:3.72/4.0 (23/161) -Advisor:[Yazhou Ren](http://smilelab.uestc.edu.cn/members/renyazhou/) **University of California,Santa Barbara(UCSB)** eschange, Computer Science Lesson in progress:Machine Learning; Computer Graphics; Computer Communication Networks ### PUBLICATION - Jiayu Guo, M.D.; Weili Zhang; **Zewen Cui**; Anran Liu; Yongmei Yan, Photonic Sensors, PHSE-D-19-00159” Dental Imaging with Near-infrared Transillumination Using Random Fiber Laser” [PDF](http://github.com/cuizewen/cuizewen.github.io/raw/master/patent.pdf) - **Zewen Cui**,Weili Zhang,Jiayu Guo, Chinese national patent ”Dental diagnosis method based on Optical fiber random laser near infrared imaging”[PDF](http://github.com/cuizewen/cuizewen.github.io/raw/master/patent.pdf) ### RESEARCH EXPERIENCE **Research on Deep Clustering Algorithm based on Self-step Learning** **Center for Artificial Intelligence and Smart Health(CAISH)，UESTC** advisor:Yazhou Ren, PhD of George Mason University * Tried to improve the original depth clustering algorithm, and the model achieved ACC=0.973±0.003, at least 14% higher than the baseline (DeepCluster=0.854). * Used deep neural network (CNN) for clustering, and applied data augmentation in pretraining stage and finetuning stage to alleviate the problem of overfitting in the depth clustering. * Incorporated adaptive step-size learning (self-paced learning) to exclude samples far away from the clustering center, and solved the problem of unapproachable edges. **Deep Learning in Computer Vision for Self-Driving Cars** **Artificial Intelligence Laboratory, Tsinghua University** advisor:Zihan huang, PhD of Tsinghua University - Combined CNN and GRU to realize automatic recognition of graphic verification code. (captcha library is used to generate the verification code randomly.) - Applied Adam optimization method to solve the non-convex optimization problem of neural network-based model on high-dimensional and large-scale data set. - Implemented CNN and GRU based with CTC loss function to obtain 99.31% accuracy on captcha dataset. **Study on Random Lasers Tooth Imaging based on Genetic AlgorithmState** **key laboratory of optical fiber sensing, UESTC** advisor:Weili Zhang,PhD of Nanyang Technological University * Designed a transmission reflection random laser bioimaging device for better effectiveness compared with the X-ray imaging commonly used in medicine. than the X-ray imaging commonly used in medicine. * Improved the existing experimental device and combining piezoelectric tube with single-channel optical fiber, where the mode number of light sources could be controlled accurately. * Added genetic algorithms for the traditional optical imaging to ensure the stable imaging effect of biological tissue in the practical application. ### AWARDS - Excellent student scholarship*2(Top 15%) - Overseas exchange scholarship: Singapore management university (3/50) - The first prize of innovation and entrepreneurship team of university students (5/71) ### INTERNSHIP **Creative Fountain Pte Ltd** - Participated in the development of WeChat terminal small program: Findapro, and mastered the development process of QT and Dreamweaver CC in the current terminal development. - Mastered the WeChat small program own a set of framework system, Javascript+CSS, mastered to call the relevant API in the project, WeChat native application effect. - Implemented tag logic and data binding with Vue, and defined new tag libraries with react instead of HTML. ### SKILLS/ INTERESTS - Software: matlab, C, python, java, tensorflow, keras - Hardware: FPGA, STM32, keil, vivado