Figure 5 - available via license: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
Content may be subject to copyright.
Visualization of the contextual centroid perception on the validation spit of the KITTI dataset. All representative points and predicted centroid are colored in gold and red, respectively. In particular, we also show the offsets of representative points inside/around the objects in red/gold. Best viewed in color.
Source publication
We study the problem of efficient object detection of 3D LiDAR point clouds. To reduce the memory and computational cost, existing point-based pipelines usually adopt task-agnostic random sampling or farthest point sampling to progressively downsample input point clouds, despite the fact that not all points are equally important to the task of obje...
Context in source publication
Context 1
... Visualization of the Contextual Centroid Perception. We also visualize the results produced by our contextual centroid perception module in Figure 5. It is clear that the downsampled point clouds at this stage are quite sparse and insufficient, which makes the centroid estimation and instance regression considerably difficult. ...
