GraVoS: Voxel Selection for 3D Point-Cloud Detection
This work addresses dataset imbalance issues in 3D point-cloud detection, offering a general solution applicable to various detectors, though it is incremental as it modifies existing scenes rather than introducing a new paradigm.
The paper tackles the challenges of 3D object detection in large scenes, such as sparsity, irregularity, and dataset imbalances, by proposing a method to select meaningful voxels for removal rather than adding objects, which improves the performance of multiple voxel-based detectors.
3D object detection within large 3D scenes is challenging not only due to the sparsity and irregularity of 3D point clouds, but also due to both the extreme foreground-background scene imbalance and class imbalance. A common approach is to add ground-truth objects from other scenes. Differently, we propose to modify the scenes by removing elements (voxels), rather than adding ones. Our approach selects the "meaningful" voxels, in a manner that addresses both types of dataset imbalance. The approach is general and can be applied to any voxel-based detector, yet the meaningfulness of a voxel is network-dependent. Our voxel selection is shown to improve the performance of several prominent 3D detection methods.