Olga Vysotska
Papers
11
Total Citations
358
H-Index
9
About
Olga Vysotska is a leading researcher in long-term robotic autonomy, specializing in visual place recognition, localization, and mapping under extreme environmental change. Her work addresses one of robotics’ most persistent challenges: enabling robots to reliably navigate environments that undergo dramatic seasonal, weather, and illumination shifts. Her most influential contribution is the development of “lazy data association” techniques for matching image sequences across substantial appearance changes, a method that has garnered 68 citations and laid the foundation for robust, long-term visual localization. She further advanced the field with multi-sequence mapping approaches (54 citations) and innovative use of publicly available maps like OpenStreetMap for global localization using compact 4-bit semantic descriptors (50 citations). Vysotska has also pioneered methods to exploit building information from open-source maps to improve SLAM, reducing the need for dedicated mapping phases. Her work spans diverse applications, from agricultural robotics—tracking growing plants over time—to traversability analysis with low-cost sensors and exploration of challenging underground environments like catacombs. With over 350 total citations, her research is essential reading for anyone working on visual navigation, place recognition, or long-term robot deployment in changing real-world settings.
Research Focus
Key Achievements
Top Papers
- 1
- 2Effective Visual Place Recognition Using Multi-Sequence Maps54 citations · 2019
- 3Global Localization on OpenStreetMap Using 4-bit Semantic Descriptors50 citations · 2019
- 4
- 5
- 6
- 7Efficient traversability analysis for mobile robots using the Kinect sensor32 citations · 2013
- 8Exploration and mapping of catacombs with mobile robots12 citations · 2013
- 9
- 10Adaptive Robust Kernels for Non-Linear Least Squares Problems3 citations · 2021