Search papers, labs, and topics across Lattice.
3
0
3
2
Radar SLAM can now achieve state-of-the-art performance via direct scan registration, eliminating the need for hand-engineered feature extraction and enabling robust localization in adverse weather.
Surprisingly, fusing a 2D imaging radar with a gyroscope achieves lidar-level odometry accuracy, unlocking a potentially cheaper and more robust alternative to lidar-based navigation.
State-of-the-art SLAM algorithms can fail to re-localize in changing seasons, as highlighted by a new multi-modal, year-long boreal forest dataset.