Depth SLAM enhanced by Machine Learning

SLAM is an active and open research problem for the robotics and computer vision communities. In this project, we propose the study of geometric information obtained from  depth sensors such as 2D LIDARs, 3D LIDARs, ToF or RGBDs to be used in SLAM, both in the front end by extracting features and in the back end for more efficient calculations [1]. In this process of obtaining features, we propose to exploit geometric features, such as planes [2] or other regular surfaces and the use of machine learning to provide new alternatives.

[1] https://github.com/MobileRoboticsSkoltech/mrob
[2] Ferrer, G. “Eigen-Factors: Plane Estimation for Multi-Frame and Time-Continuous Point Cloud Alignment.” IROS. 2019.


  • Candidates should hold or expect an MSc (or an equivalent) degree in the following or closely related disciplines:
    – Robotics
    – Data Science/Computer Science
    – Engineering
  • Basic knowledge in Programming
  • Programming skills: Cpp and Python. ROS is a plus
  • High interest in Robot Perception and ML
  • Previous experience in Robotics, ML
  • Motivation to complete a PhD within 4 years