PillarNeSt: Embracing Backbone Scaling and Pretraining for Pillar-based 3D Object Detection

Open in new window