Introduction

Welcome to my Pokeball! My name is Menghe. Currently, I am pursuing my Ph.D. degree in the Center of Visual Computing, UCSD, supervised by Dr.Jürgen P. Schulze.

My research interests focus on investigations to improve Augmented Reality with better computational imaging, user interaction and localization.

Experiences

  • Research Assistant@ Samsung

    2018/06 - Present

    Augmented Reality (AR) Rendering: investigate methods for augmented reality, in particular mixed reality methods that blend rendering of real and virtual environments.

  • Student Developer@ Google Summer of Code, NRNB

    2017/06 - 2017/09

    GPU-accelerated Rendering for Cytoscpe: design and develop the prototype of 2d network graph renderer, which takes advantage of high speed GPU calculation.

  • Develop Engineer@ Cocos2d-x Engine Group

    2015/06 - 2015/12

    Developed and maintained Cocos2d-x game engine and Cocos Studio. Wrote parts of official programmers’ guide and reviewed contributions from external developers. Refactored UI modules and improve performance of WebGL rendering in both 2D and 3D.

  • Research Assistant@ Xiamen University

    2014/01 - 2015/06

    The Attention Training System based on Neuro-feedback: design and develop the prototype of attention training system using Unity3D based on real-time analysis of EEG signals.

  • Research Assistant@ Xiamen University

    2012/09 - 2013/12

    Multiple Underwater Simulation Platform of Intelligent Robot: build a simulation platform of varied underwater robot controls.

Projects

Augmented Reality Assisted Laparoscopic Surgery with Magnetic Resonance Imaging (MRI)

CSE 190/291: Human-Centered Computing for Health Final Project

Mathematics behind Rigidbody Simulation

Implementations of some physics simulations that combines different components: physics, collision detection, rendering, user interaction, etc.

Interior Design with Automated Furniture Layout

Augment an existing interior layout with a new layout design by automated layout algorithms and user interaction.

LSTM Action Recognition

Design and implement LSTM auto encoder and future prediction system to assist action recognition.

Practival Black-Box Attacks against Deep Learning Systems using Adversarial Examples

Generate adversarial examples to fool Deep Neural Networks without knowing the details of networks

High Quality Rendering in OpenGL

Investigation on several different shadow mapping and do environment mapping.

Games and 3D Scene Rendering

Games includes VR games developed on Oculus and Unity3D(3D Tetris, flying game...) and common 3D games like car racer and so on.

Contact Me

Address
9500 Gilman Drive • La Jolla, 92093 • USA
Email
mez071 AT eng.ucsd.edu