German Outdoor and Offroad Dataset
The German Outdoor and Offroad Dataset (GOOSE), along with related datasets like TartanDrive and Oxford Offroad Radar, provides a crucial resource for advancing autonomous navigation in challenging environments. Current research focuses on improving perception models using various sensor modalities (LiDAR, radar, RGB cameras) and developing robust algorithms for semantic segmentation, traversability estimation, and place recognition, often employing deep learning architectures like neural networks. These datasets and associated research efforts are significantly impacting the development of safer and more reliable autonomous systems for off-road applications, including robotics and autonomous vehicles.
Papers
OORD: The Oxford Offroad Radar Dataset
Matthew Gadd, Daniele De Martini, Oliver Bartlett, Paul Murcutt, Matt Towlson, Matthew Widojo, Valentina Muşat, Luke Robinson, Efimia Panagiotaki, Georgi Pramatarov, Marc Alexander Kühn, Letizia Marchegiani, Paul Newman, Lars Kunze
UFO: Uncertainty-aware LiDAR-image Fusion for Off-road Semantic Terrain Map Estimation
Ohn Kim, Junwon Seo, Seongyong Ahn, Chong Hui Kim