Event
VPR-Space Extraterrestrial Surface Visual Place Recognition Benchmark Released
1 reports · 1 sources
Richard Fechner and Peer Neubert have released the VPR-Space benchmark for evaluating visual place recognition (VPR) capabilities in extraterrestrial surface scenarios such as the Moon and Mars. The benchmark is built on existing data from the Space Analog Domain (SAD) and comprises 16 database-query datasets, covering challenges such as environmental, viewpoint, and illumination variations. The authors state that a pre-release version of the dataset is available on GitHub.
Report timeline
- VPR-Space: A Benchmark for Visual Place Recognition in Space Analog Domains
The paper proposes the VPR-Space benchmark to evaluate visual place recognition (VPR) capabilities in extraterrestrial surface scenarios such as the Moon and Mars. The authors built 16 database-query datasets based on existing data from Space Analog Domains (SAD), covering challenges including environment, viewpoint change, and illumination conditions.
arXiv cs.RO ·