Assistant Machine

Zhizuo(Victor) Zhang

**** ***** *****, *** ****, 92122 - San Diego, CA - United States

Æ +1-858-***-**** or +86-136-****-****)

Q *******@****.*** or ***********@*****.***

LinkedIn: https://www.linkedin.com/in/zhang-zhizuo-14740514b; Github:https://github.com/MachineryZ

Language and Skills

+ English: Fluent

+ Chinese: Native Speaker

+ Languages: C++, C, Python, HTML, MATLAB, Java

+ Skills: Machine Learning, Deep Learning

+ Tools: Windows, Linux/Unix, Pytorch, Keras, Tensorflow, Vim, Git Education Background

+

Bachelor of Science: Intelligence Science And Technology Department Peking University Sep 2015 - June 2019 Beijing, China

- Mathematical Courses: Mathematical Analysis(I)(II)(III), Linear Algebra(B), Set Theory and Graph Theory, Probability Theory and Statistics(A), Algebraic Structure and Combinatorial Mathematics, Numerical Methods(B), Mathematical Logistics

- Computer Science Courses: Practice of programming in C&C++, Basic Experiments of Electrocircuit, Biometrics, Data Structure and Algorithm(A), Introduction to Intelligent Technology, Introduction to Computer Systems, Design of Digital Logic, Algorithm Design and Analysis, Advanced Machine Learning, Introduction to Pattern Recognition, Introduction to Artificial Intelligence, Signals and Systems, Deep Learning

+

Master of Science: Machine Learning & Data Science, ECE UCSD Sep 2019 - Expected Apr 2021 San Diego, CA, United States

- Mathematical Courses: Linear Algebra, Convex Optimization, Probability & Statistics, Random Processes

- Computer Science Courses: Statistical Learning (I)(II)(III), Optimization & Acceleration, Computer Vision (III),

Experience

+

Research Assistant (Compression on MRI image) Zero Lab Sep 2017 - Aug 2018 Professor Zhouchen Lin (PKU)

- Research on basic idea of deep learning and neural network

- Use Clique-Net("Convolutional Neural Networks with Alternately Updated Clique") to compress and decompress MR image

- Replace RNN block("Full Resolution Image Compression with Recurrent Neural Networks") into Clique Block

- Modified numerous parameters of Clique RNN to optimize result

- Compression rate are 1/8, average SSIM on dataset is 0.9976

- First time to get involved into project about deep learning and neural network 1/2

+

Summer Research Assistant (EM algorithm on CECT) Carnegie Mellon University Jul 2018 - Sep 2018 Professor Min Xu (CMU)

- Research on EM algorithms, Gaussian Mixture Model and its advance weighted GMM

- Derive updating process of EM algorithms from "Averaging of Electron SubTomograms and Random conical Tilt Reconstructions through Likelihood Optimization"

- Reconsider missing edge effect and unobserved data through original EM algorithms

- Emulate previous EM algorithm’s and weighted EM algorithm’s convergence rate

- Transfer weighted EM algorithms to latter research co-workers

+

Research Internship (Segmentation on pulmonary CT image) SenseTime Company Dec 2018 - Apr 2019 PhD. Ning Huang

- Work in ’SenseCare’ department

- Work closely with PhD. Ning Huang to deal with pulmonary CT image segmentation problem

- To separate the trachea from the background

- Use modified V-Net model called DenseV-Net to finish segmentation

- Come up with using Focal Loss combined with CE Loss to train the model

- Use DenseV-Net to separate heart, trachea, iota and esophagus from SegThor dataset

- Dice results are 93.2%, 87.6%, 86.6%, 76.3%

+

Research Assistant (Robotics) Sulab

Nov 2019 - Now Professor Hao Su (UCSD)

- Research on sample based method, RRT, RRT*, PRM.

- Research on “Lego: leveraging experience in roadmap generation for sampling-based planning” and

"MPNet"

- Apply segmentation and detection backbone network to find landmark points.

- Perform better than Lego on 2D environments’ experiment.

- Still in progress

Course Project

+ Research on Support Vector Machine

+ Stopwatch & Calculator on FPGA

+ Reproduce some topics in Computer Vision: face detection, structure from motion, human pose estima- tion.2

+ Summary on Variants of GAN

+ Summary on Quadratically constrained quadratic program and its relaxation

+ Research on GNN & GCN

Rewards and Examination

+ Toefl: 100

+ GRE: 324+3.5

+ Chinese Physics Olympaid (CPhO) 2014: Silver Medal

+ ACM: Bronze Medal of Beijing District.

2/2

Contact this candidate