Previous and ongoing projects

LiDAR based Place Recognition

November 10, 2018

Projects, Carnegie Mellon University, RI, Pittsburgh, USA

For the appearances changes under different weather or season conditions, we proposed a Weather condition invariant unsupervised feature extraction method. By using a conditional cycleGAN method, we achieve to split the weather conditions from the geometry features, without losing the geometry detail at the same time. The extracted weather condition invariant features could make better matching under different weather conditions. The below figure shows the reconstruction result with the extracted geometry features under different conditions. The experiment results could be seen from the below website.

Conditional Invariant Feature Learning for VPR task

April 09, 2018

Projects, Carnegie Mellon University, RI, Pittsburgh, USA

For the appearances changes under different weather or season conditions, we proposed a Weather condition invariant unsupervised feature extraction method. By using a conditional cycleGAN method, we achieve to split the weather conditions from the geometry features, without losing the geometry detail at the same time. The extracted weather condition invariant features could make better matching under different weather conditions. The below figure shows the reconstruction result with the extracted geometry features under different conditions. The experiment results could be seen from the below website.

A Fusiong Framework for LiDAR-based Place Recognition Method

September 10, 2017

Projects, Carnegie Mellon University, RI, Pittsburgh, USA

Visual place recognition is a challenging problem due to the vast range of ways in which the appearance of real-world places varies, the major problems come from:

  • Illumination changes, scenes under light conditions may look totally unfamiliar.
  • Appearance changes, whether or season caused appearance change, such as trees under summer and winter will show a different outlook.
  • Viewpoints changes, the same accurate pose may never occur twice, the extract features need to be viewpoints invariant, however, capture the description for a specific scene at the same time.

Field Autonomous Driving

September 01, 2016

Projects, University of Chinese Academy of Sciences, Shenyang, China

The mission is aim to enable the Autonomous Car travel about 15~20 km in the unstructured terrain area. Because there is no obvious road shape in the terrain area, efficient road shape estimation method need to be taken; On the other side, trees and tunnel may also shield the GPS signal, so efficient LiDAR-Odometry and Visual-SLAM method needs to be added into the localization structure.

DJI Developer Challenge 2016

June 10, 2016

Projects, University of Chinese Academy of Sciences, Shenyang, China

The mission is twofold, first the UAV should take the vertical taking off and landing control on a dynamic car (which is driving at 5m/s on average); Secondly, the UAV system should recognize the targets in the unstructured environment.

Crossing Area Multi-Robot Cooperation System 2016

April 15, 2015

Projects, University of Chinese Academy of Sciences, Shenyang, China

We design a multi-robots cooperation system, which contains a UAV (Unmanned Aerial Vehicle) and a UGV (Unmanned Ground Vehicle). UAV system is good at long term manipulation, however, it’s weak in global navigation; UGV is flexible in global navigation but weak in endurance ability. Multi robots system could make up the shortages of different robots. However, the key problem is that, the scan from different robots may have a totally different resolution, scale, format etc. How to make the accurate cooperate localization between Large scale-Coarse point clouds with Small scale-Dense point clouds.