HOMEPAGE

Red and black quadcopter drone flying down a marble-floored hallway toward a staircase

Overview

The objective of the Multiple Autonomous Robotic Systems (MARS) Laboratory is to promote basic research and education in robotics and computer vision with special emphasis on estimation and control of autonomous ground, aerial, and space exploration vehicles.

Researchers in our group have a diverse background in Computer Science and Engineering, Electrical Engineering, Aerospace Engineering, Industrial Engineering, and Physics.

The research interests of our group are on resource-aware estimation and control, including:

  • Vision/Laser-aided inertial navigation systems (VINS, RGBD-INS, LINS)
  • Large-scale 3D localization and mapping on cell phones and wearable computers
  • Multi-robot/sensor localization, mapping, and navigation
  • Active sensing for reconfigurable networks of sensors
  • Optimal information selection and fusion
  • Mobile manipulation
  • Human-robot cooperation

Research Highlights

White Google Glass headset resting on a gray surface
Red and black Parrot Bebop quadcopter with four propellers
Stack of overlapping photos of the Golden Gate Bridge
Smartphone held up in a hallway, its screen overlaying tracked feature points and a pink camera marker
Person holding a tablet in a library aisle, its screen showing tracked 3D feature points

 Incremental BLS

Zoomed-in floor plan with a blue arrow showing the estimated position and heading in a corridor

 Map-based Localization

Two cameras, shown as blue triangles, observe green stars, with a crossed-out equals sign between the stars

C-KLAM

Semi-dense 3D point cloud of a staircase with a balustrade, rendered on a dark blue background

 Semi-Dense Mapping

Diagram of epipolar geometry: a point X seen by two cameras, with a red epipolar line in the right image

 Stereo VINS

Video

The MARS Lab Youtube Channel

Visualizations

Building floor plan with a pair of red footprints marking the user’s current location

Latests News

Nov. 8 2019: "Visual Inertial Systems: A Tutorial" (pdf)

May. 10 2016: Dr. Nikolas Trawny, UMN CS PhD'10 and former MARS Lab member, now at JPL, NASA has been appointed as the Lead for the Europa Lander Intelligent Landing System. This will be an amazing journey! Congratulations!

Apr. 25 2016: Dr. Luis Carrilo-Arce, MARS Lab Postdoctoral Fellow, joins Google's Project Tango. Congratulations!

Dec. 16 2015: Zhengqi Li has been selected for honorable mention for the Computing Research Association's (CRA) Outstanding Undergraduate Male Researcher Award for PhD-granting institutions 2016. Congratulations!

Nov. 24 2015: Prof. Stergios Roumeliotis is elevated to IEEE Fellow (2016) for his contributions to Visual-Inertial Navigation and Cooperative Localization

Apr. 27 2015: C. Guo receives the UMN Doctoral Dissertation Fellowship Award. Congratulations!

Mar. 18 2015: Congratulations to Nikolas Trawny, former MARS Lab member, and his colleagues at JPL, NASA for the successful test flights of Mars Landing Technologies [JPL News]

More News

Recent Publications

T. Ke, and S. I. Roumeliotis, "An Efficient Algebraic Solution to the Perspective-Three-Point Problem," (CVPR'17).

M. K. Paul, K. Wu, J. A. Hesch, E. D. Nerurkar, and S. I. Roumeliotis, "A Comparative Analysis of Tightly-coupled Monocular, Binocular, and Stereo VINS," (ICRA'17).

K. J. Wu, C. X. Guo, G. Georgiou, and S. I. Roumeliotis, "VINS on wheels," (ICRA'17).

R. C. DuToit, J. A. Hesch, E. D. Nerurkar, and S. I. Roumeliotis, "Consistent map-based 3D localization on mobile devices," (ICRA'17).

More Publications