Self-supervised Lidar place recognition in overhead imagery using unpaired data

Conference item

As much as place recognition is crucial for navigation, mapping and collecting training ground truth, namely sensor data pairs across different locations, are costly and time-consuming. This paper tackles these by learning lidar place recognition on public overhead imagery and in a self-supervised fashion, with no need for paired lidar and overhead imagery data. We learn the cross-modal data comparison between lidar and overhead imagery with a multi-step framework. First, images are transformed into synthetic lidar data and a latent projection is learned. Next, we discover pseudo pairs of lidar and satellite data from unpaired and asynchronous sequences, and use them for training a final embedding space projection in a cross-modality place recognition framework. We train and test our approach on real data from various environments and show performances approaching a supervised method using paired data.

Authors: Tang, TY; De Martini, D; Newman, PM

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Robotics: Science and Systems Foundation
• Host title: Proceedings of Robotics: Science and Systems 2023
• Article number: 98
• Series: Robotics: Science and Systems Proceedings
• Series number: XIX
• Publication date: 2023-10-07
• Acceptance date: 2023-04-22
• Event title: 19th Robotics: Science and Systems (RSS 2023)
• Event location: Daegu, Republic of Korea
• Event website: https://roboticsconference.org/2023/
• Event start date: 2023-07-10
• Event end date: 2023-07-14
• DOI: 10.15607/rss.2023.xix.098
• ISSN: 2330-765X
• ISBN: 9780992374792
• Language: English