A Short Bio

My research focuses on developing perception algorithms for robotics, such as visual SLAM, visual inertia odometry, multi-modal sensor fusion with GPS/odometer and etc,. In the recent years after graduation, I also developed coverage planning algorithms on both robotic arms and mobile platforms for industrial application. Except from developing softwares and algorithms, I enjoy building different kinds of hardware platforms for toy robotic applications during my part-time. I married with Dr. Liye Sun who is also doing research work in the field of robotics and computer vision and have a 3-years old cute baby girl.

Education Background

I spent most of my childhood time in Lu’an, Anhui with nothing in particular. When I was at the age of thirteen, I was fortunate enough to visit Jiuquan Satellite Launch Center located on the Gobi Desert. This later became the most important motivation when I selected astraunautics as the major of the my master study. Till today, I still believe that rocket science especially Mars rovers such as Curiosity is one the most complicated and advanced robotic system on this planet. I enjoyed reading science fictions magazines during my young age and Science Fiction World is my favorite.

I completed my bachelor degree at Northwestern Polytechnical University, located in Xi’an Shaanxi from 2006 to 2010. I was able to learn different subjects as a student at Honor College (formly known as Education Experimental School), from electronic engineering to automation, from math to computer science. I also met my wife during the three beautiful years at Honor Colledge. I moved School of Astraunautics in the last year of my bachelor work and completed my Master degree there. I was working on state estimation of cooperative or even hostile spacecrafts using computer vision and orbital mechanics. Since then, I was getting to know visual SLAM, multiple view geometry and other tools which were used in the next 10 years till now. I have earned multiple awards during my master work and got tuition waiver throughout my two and a half years of master study.

From the December of 2012, I started my Ph.D. work at University of Technolody, Sydney, supervised by Prof. Gamini Dissanayake. Spending fours years pursuing a Ph.D degree is nothing difficult and painful for me, oppositely, fours years at Center for Autonomous System is the most impressive and memberable experience during my whole life. I have the opportunity to work with Prof. Shoudong Huang, Prof. Dikai Liu, Dr. Ravindra Ranasinghe and Dr. Liang Zhao, and met so many good friends such as Dr. Teng Zhang, Dr. Kasra Khosoussi, Dr. Lakshitha (Laki) Dantanarayana, DR. Raphael Falque and on and on. My wife also did her Ph.D work with me at the same lab. I completed my Ph.D work at the Decemeber of 2016 and received my Ph.D. degree in 2017.

Research Work

The thesis of my Ph.D work is Active Perpcetion for Robot Manipulation which covers three topics:

  • Instance detection and pose estimation which provides a semantic understanding of the current scenario and is used as a prior knowledge for the next planning step
  • Next-Best-View planning finds the optimal trajectory in a next few steps which maximizes the information gain and also minimizes the state estimate uncertainty
  • Multi-model sensor fusion estimates the robot pose while the robot moves in the environment by fusing Lidar, vision, IMU and encoder measurments

The ultimate goal of my Ph.D work is to plan the optimal trajectory of a mobile robot to find the targeting objects in indoor environments and estimate the pose of the robot simultaneously. It is one of the most fundamental task for service robots when the robots is asked to find and pick up a certain item within the room.

The publication list during the my Ph.D work is below and it is the best way to get a quick understanding of the my research work or visit my Google Scholar page.

Publication List

  • Convergence and consistency analysis for a 3-D invariant-EKF SLAM at ICRA 2017/RAL
  • An Invariant-EKF VINS Algorithm for Improving Consistency at IROS 2017
  • Active recognition and pose estimation of household objects in clutter at ICRA 2015
  • RISAS: A Novel Rotation, Illumination, Scale Invariant Appearance and Shape Feature at ICRA 2017
  • Gyro-aided camera-odometer online calibration and localizatio at ACC 2017
  • Planar scan matching using incident angle at IROS 2017
  • A fast pipeline for textured object recognition in clutter using an RGB-D sensor at ICARCV 2014
  • Active Object Detection and Pose Estimation in General Belief Space at RSS 2016 Workshop

Industrial Engagement