CONTACT ME
E-mail: ziyezhou.jerry@gmail.com
Phone: 0086-15201521817
EDUCATION
09/2010 – Present
Tsinghua University, Beijing, China
B.S. in Automation, College of Information and Technology.
Overall GPA: 90.55/100 , Ranking: 6/151.
AWARDS AND HONORS
- Oct. 2011 Tsinghua’s Friend-EMC Scholarship for comprehensive excellence (awarded top 5% students)
- Oct. 2012 Tsinghua’s Friend-SuZhou industrial estate Scholarship for comprehensive excellence (awarded top 5% students)
- Sep.2012 Second Prize in the National Undergraduate Physics Competition
- Oct.2013 Scholarship for academic excellence (awarded top 10% students)
RESEARCH EXPERIENCES
Aug.2013 – present
Patch-Based Plenoptic Image Manipulation
advised by Professor Shi-Min Hu Graphics & Geometric Computing Group, Tsinghua University
- Focused on editing the light field captured by the lenselet devices and synthesize new light field data
- Proposed a fast way to recover the depth information based on plenoptic imaging technique
- Given the definition of patch in the plenoptic image and the similarity measurement
- Paper intended to submit to SIGGRAPH 2014
July. 2013 – Sep. 2013
Fast Portrait Retrieval based on Clothing Extraction Program
advised by Professor Shi-min Hu Graphics & Geometric Computing Group, Tsinghua University
- Present a fast portrait image retrieval method based on effective clothing extraction algorithm, the result of which outperforms the tradition methods based on image similarity in portrait retrieval task.
- Proposed a method to extract the clothing region in portrait images without training process, which use multi-label graph-cut method to optimize the labels for the over-segmented sub-regions on each images, and used the features of the clothes region to enable an efficient portrait retrieval framework.
- Built a human portrait library to test the performance of previous method, which contain more than 30,000 pictures taken manually or downloaded from the Flickr.
-
Paper submitted to Frontiers of Computer Science (FCS)
July. 2012 – July. 2013
Intelligent Image Search System
advised by Professor Shi-min Hu Graphics & Geometric Computing Group, Tsinghua University
- Focused on searching the valid (with person, with car, etc.) pictures from numerous surveillance images and then extract the information users want.
- Proposed one multi-patches federated searching method based on the implementation of template match method, which greatly improve the searching performance.
July. 2013 – Oct. 2013
Research intern
Megvii Inc, Beijing China
- Proposed one face-region segmentation method on the basis of grab-cut and embedded it into Megvii’s commercial face platform, which greatly improve the performance of their face warping application. This face platform is under further development to show on the ICCV 2013.
- Involved in Megvii’s research team and cooperated with other researchers to improve Megvii’s own face recognition cloud platform---Face++ based on the up to date face feature point detection algorithm, and the key algorithm has already got the patent from China Patent Office.
Sept. 2011 – June. 2012
Student Research Training Program
advised by Professor Qian-Chuan Zhao Center for Intelligent and Networked Systems, Tsinghua University
- Proposed an efficient aggregate method to turn the stage with lots of machines connected in parallel/in series into an equivalent single machine, which performed 2 times faster on the same dataset than the previous model.
- Developed a crew scheduling program based on the previous working, which is adopted by the China Railway Customer Call Service Center.
SELETED PUBLICATION
Fast Bust Portrait Retrieval based on Clothing Extraction, Zi-Ye Zhou, Fang-Lue Zhang, Shi-Min Hu, submitted to Frontiers of Computer Science (FCS)
TECHNICAL SKILLS
- Programming language: C/C++, Java, Python, C#, Assemble Language and Verilog HDL
- Programming library: OpenCV, OpenGL, Qt
- Programming software: MATLAB, Octave, Quartus II, PSPICE, Multisim, and Proteus
STANDARD TEST:
- TOEFL: Reading:28 Listening:28 Speaking:23 Writing:27 Total Score:106
- GRE: Vebal:159 Quantity:166 Total Score:325 AW:4.0