Hi! My name is Seva. Here is some information about my life, hobbies, and research projects.

Donate to the Ukrainian army:

“I would rather have questions that can’t be answered than answers that can’t be questioned” - Richard Feynman

Currently, I am doing a PhD in Robotics Navigation at the COMROB lab at the Czech Technical University in Prague :)

Personal

Hobbies

Other

Professional

Education

Work Experience

CV for a research position (Updated 07.2023)

CV for a research position (Updated 07.2023)

Current Research

Currently, I am focused on detecting perceptual degeneracies in sensor measurements and extending this concept to achieve robust localization through the tight fusion of sensors of different natures. Specifically, the sensors are 3D LiDAR, Camera and IMU.

I am interested in applying mathematical theories, such as Lie Groups, to describe the general degeneracy of the scan alignment problem, and then enhancing it with supplementary sensors.

Perceptually degenerated scenario: aligning a part of the cylinder to a cylinder is ambiguous.

Recent news

• Upcoming ECMR 2025 Workshop on Resilient SLAM. Features: Results of SLAM Benchmark, papers on relevant topics, and invited speakers.

• March 2024: Submitted an IROS paper, “LiDAR-Visual-Inertial tightly-coupled odometry with adaptive learnable fusion weights”. UPD July 2024: the paper is accepted.

LiDAR and Visual features which are jointly optimized. green and red: visual features, black: LiDAR features, grep: LiDAR map.

• May 2023: Submitted my master’s thesis “Failure-resilient Graph-based SLAM for Autonomous Robotic Exploration in GNSS-Denied Environments” - details, videos UPD: the thesis is defended.

• May 2023: Submitted an **ECMR paper** “Graph-based LiDAR-Inertial SLAM Enhanced by Loosely-Coupled Visual Odometry” - video. UPD: The paper is published.

Contacts

Email

[email protected] / [email protected]

LinkedIn

Facebook

Seva Hulchuck

GitHub

sevahul - Overview

YouTube

Research videos