Torsten Sattler
Torsten Sattler works on visual localisation, mapping, and reconstruction. He worked at ETH Zurich before joining Chalmers in January 2019 and CIIRC at Czech Technical University in July 2020. He co-authored work on benchmarking outdoor visual localisation in changing conditions, first submitted in 2017 and published at CVPR 2018. The benchmark examines localisation under changes such as weather and lighting. His research also covers neural scene representations and learned visual features.
- Works
- 267
- Citations
- 15,068
- h-index
- 57
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
Before 2019
Geometry and vision research
Sattler earned a doctorate at RWTH Aachen and spent five years in postdoctoral and senior research roles at ETH Zurich. His research combined 3D computer vision with machine learning.
-
2017–2018
Localisation in changing conditions
He co-authored Benchmarking 6DOF Outdoor Visual Localization in Changing Conditions. The manuscript was first submitted in 2017 and the work appeared at CVPR 2018. It provides benchmarks for camera localisation when illumination, weather, or seasonal conditions change.
-
2019–2020
Chalmers and Prague
He joined Chalmers in January 2019 as a tenured associate professor. He moved to CIIRC at Czech Technical University in Prague in July 2020.
-
As of October 2026
Scene understanding and geometry
His research covers mapping, reconstruction, learned local features and neural scene representations. He aims to make localisation and mapping more reliable by using higher-level scene understanding. His work has applications in autonomous robots and augmented reality.
Sources and credits
Photo: Roman Sejkot; source: Torsten Sattler personal profile. License: Not stated by source. Source: tsattler.github.io. Cropped to a square and resized. Ask us to remove or replace this photo.