Paper Info
Pointillism: Accurate 3D Bounding Box Estimation with Multi-Radars
K Bansal, K Rungta, S Zhu, D Bharadia
Association for Computing Machinery, 340–353, 2020.
Reviews
From
yassendobrev, rated
5/5.
Jan. 7, 2021.
An interesting paper, where they use spatial diversity (2 incoherent radars spaced 1.5 m apart) and a pointnet-based neural net to infer bounding boxes from radar point clouds. They also develop a CPPC (cross potential point cloud) metric which describes the correlation between the data of the 2 radars to try to distinguish between real targets and noise.
Pointillism: Accurate 3D Bounding Box Estimation with Multi-Radars
K Bansal, K Rungta, S Zhu, D Bharadia
Association for Computing Machinery, 340–353, 2020.
Reviews
From
yassendobrev, rated
5/5.
Jan. 7, 2021.
An interesting paper, where they use spatial diversity (2 incoherent radars spaced 1.5 m apart) and a pointnet-based neural net to infer bounding boxes from radar point clouds. They also develop a CPPC (cross potential point cloud) metric which describes the correlation between the data of the 2 radars to try to distinguish between real targets and noise.