The AI x Mobility Lab (Prof. Hwasoo Yeo, KAIST Smart City Research Center), in collaboration with the LUTS Lab at EPFL, has publicly released the Songdo Traffic and Songdo Vision datasets, derived from a large-scale multi-drone experiment conducted in the Songdo International District on October 3–7, 2022. Using an advanced computer vision algorithm that extracts geo-referenced trajectories from drone videos, the Songdo Traffic dataset provides approximately 700,000 geo-referenced vehicle trajectories, and the Songdo Vision dataset provides 300,000 annotated vehicle bounding boxes captured across 20 intersections.
Songdo Traffic Dataset: https://doi.org/10.5281/zenodo.13828384
Songdo Vision Dataset: https://doi.org/10.5281/zenodo.13828408
Source: KAIST AI x Mobility Lab – News
Comments
Post a Comment