Oct. 2016 - Present
Research Assistant, SECLAB Lehigh University
Deterministic Browser Project
- An execution time deterministic browser based on Firefox, which could defend time channel attack. Website: www.deterfox.com
M.S. Degree in Computer Science • Sept. 2016 - Jun. 2018
GPA: 4.0 / 4.0
Courses: Data Mining, Advanced Algorithm, Search Engine, Advanced OS, Advanced Programming
B.E. Degree in Software Engineering • Sept. 2012 - Jun. 2016
B.E. Degree in Financial Engineering • Sept. 2013 - Jun. 2016
GPA: 3.6 / 4.0
Research Assistant, SECLAB Lehigh University
Deterministic Browser Project
Research Assistant, Nankai University
Indoor Navigation System
Research Assistant, Nankai University
Fine-grain Leaf Recognition Project
Web Developer, Chinasoft International
Online Shopping Website
This project is aimed to solve the problem of finding your car in large underground parking where GPS signal can't reach. We use iBeacon, which could broadcast Bluetooth signal, to locate users. This research use a heuristic algorithm (Simulated Annealing) to get the optimized iBeacon arrangement, which consider both location precision and total cost. Besides the arrangement research, we build a complete navigation system including an Android app, a server, a database saving user and map data and an import tool that help designer to import map data from CAD (Computer Assist Design) map.
Besides the original navigation function, our app support position marking and position sharing functions, which could provide car finding and user finding services. To achieve better user experience, we use OpenGL to present the 3D map in Android app.
Introduction
This project is aimed to retrieve plant information based on its leaf picture. In the research part, we implied 37 common graphic features of previous research. These features is about leaf's contour, content and texture. After that, we use Random Forest algorithm to evaluate all features and pick out the most efficient features. Based on research result, we train a SVM model for leaf recognition. The experiment is based on 8000+ sample pictures from 270 species of plants.
For engineer part, we save the trained model in an Android app using OpenCV4Android so that the app could extract graphic features and get the predict result locally. The app support loading picture from both camera and file system. And we built a MySQL database where the app could retrieve plant information.
This project is aimed to retrieve plant information based on its leaf picture. In the research part, we implied 37 common graphic features of previous research. These features is about leaf's contour, content and texture. After that, we use Random Forest algorithm to evaluate all features and pick out the most efficient features. Based on research result, we train a SVM model for leaf recognition. The experiment is based on 8000+ sample pictures from 270 species of plants.
For engineer part, we save the trained model in an Android app using OpenCV4Android so that the app could extract graphic features and get the predict result locally. The app support loading picture from both camera and file system. And we built a MySQL database where the app could retrieve plant information.