I am an ECE Master student advised by Patricio A. Vela at Institute for Robotics and Intelligent Machines, Georgia Tech. I am a practical roboticist, interested in combining robotics and computer vision together to solve the challenges in the real world. For my Master thesis, I am currently extending my previous object grasping work in two ways. The first is incorporating 6-DoF Object Pose Estimation to propose better grasp candidates. The second is adopting Domain Adaptation method to narrow the gap between the simulation and the reality. [View my CV] (Updated in Jan. 2020).

Before coming to Georgia Tech, I received B.E. in Automation from Southeast University (SEU) in 2018. During that period, I was advised by Wenze Shao and Yangang Wang. I also interned at Applied Nonlinear Control Lab, University of Alberta, Edmonton, Canada, in Fall 2017, working with Alan Lynch. My previous research experiences include (a) Underwater Robot System Design, (b) Motion Image Deblurring, (c) Visual Servoing of Unmanned Vehicle Systems, and (d) Object Grasping via Primitive Shapes.


  • [Jan 2020] One paper accepted to  ICRA 2020  about our novel work on Using Synthetic Data and Deep Networks to Recognize Primitive Shapes for Object Grasping


  • Robotics
  • Computer Vision
  • Image Processing


  • M.S. in Electrical and Computer Engineering, 2020(Expected)

    Georgia Institute of Technology

  • B.E. in Automation, 2018

    Southeast University


Quickly discover relevant content by filtering publications.
Using Synthetic Data and Deep Networks to Recognize Primitive Shapes for Object Grasping
A new variational approach to deblurring low-resolution images
Nonparametric Blind Super-Resolution Using Adaptive Heavy-Tailed Priors
Blind Deblurring Using Discriminative Image Smoothing
Robust Blind Deconvolution Using RelativeTotal Variation as a Regularization Penalty



Object Grasping via Primitive Shapes - Research Assistant

Intelligent Vision and Automation Laboratory, Georgia Institute of Technology

Advisor : Patricio A. Vela
Feb 2019 - Present Atlanta, Georgia, USA

  • Developed an automated strategy to rapidly generate data in the simulation
  • Built up a deep network, segmentation-based pipeline for primitive shapes decomposition
  • Experimental testing and evaluation of grasping accuracy using a 7-DoF robotic arm


Ear Keypoint Detection - Research Assistant

Southeast University

Advisor : Yangang Wang
Mar 2018 - Jun 2018 Nanjing, Jiangsu, China

  • Designed a cascaded three-stage deep network for real-time ear keypoint detection
  • Collected a large scale dataset for ear keypoint detection with keypoint annotation, occlusion, and accessory information


Visual Servoing of Unmanned Vehicle Systems - Research Assistant

Applied Nonlinear Controls Laboratory, University of Alberta

Advisor : Alan Lynch
Sep 2017 - Dec 2017 Edmonton, Alberta, Canada

  • Implemented an improved image-based visual servoing algorithm on a self-designed quadrotor
  • Secondary development of PX4 firmware, an open-source flight control system
  • Environmental calibration based on the analysis of Vicon motion capture system data


Image Motion Deblurring - Research Assistant

National Engineering Research Center of Communications and Networking, Nanjing University of Posts and Telecommunications

Advisor : Wenze Shao
Oct 2016 - Mar 2018 Nanjing, Jiangsu, China

  • Introduced relative total variation as a regularization penalty for single image blind motion deblurring
  • Adopted the Operator Splitting and the Augmented Lagrangian method to solve the equations
  • Extensive comparisons of the state-of-the-art algorithms with different evaluation metrics


Underwater Robot Design - Research Assistant

National Undergraduate Innovation Research Program, Southeast University

Advisor : Qiqi Liu
May 2016 - Mar 2017 Nanjing, Jiangsu, China

  • Built an underwater robot from scratch with a lightweight design and multiple embedded systems
  • Applied MJPG-streamer with a Tornado server framework on Raspberry Pi to support real-time data transmission

Autonomous Lane Tracking for Car-like Robot - Group Leader

National Undergraduate Innovation Research Program, Southeast University

Oct 2015 - Jun 2016 Nanjing, Jiangsu, China

  • Developed a rule-based pattern recognition algorithm for real-time line tracking on Raspberry Pi
  • Applied the incremental PID algorithm to control steering motors on Arduino