ONCE
One million LiDAR scenes for self-supervised 3D perception. · By Huawei Noah's Ark
ONCE offers a million LiDAR frames and seven million camera images from Chinese roads, with a small labeled subset, to benchmark semi- and self-supervised 3D object detection.
Description
ONCE offers a million LiDAR frames and seven million camera images from Chinese roads, with a small labeled subset, to benchmark semi- and self-supervised 3D object detection.