Skip to content

Latest commit

 

History

History
43 lines (26 loc) · 2.53 KB

README.md

File metadata and controls

43 lines (26 loc) · 2.53 KB

MODEL ZOO

Common settings and notes

  • The experiments are run with PyTorch 1.1, CUDA 10.0, and CUDNN 7.5.
  • The training is conducted on 4 V100 GPUs in a DGX server.
  • Testing times are measured on a TITAN RTX GPU with batch size 1.

nuScenes 3D Detection

We provide training / validation configurations, logs, pretrained models, and prediction files for all models in the paper

VoxelNet

Model Validation MAP Validation NDS Link
centerpoint_voxel_1440 59.6 66.8 URL

Please refer to LINK for centerpoint detection predicitons on nuScenes train/val/test sets.

VoxelNet(depreacted)

These results are obtained before the sync bn bug fix + z axis augmentation .

Model FPS Validation MAP Validation NDS Link
centerpoint_voxel_1024 16 56.4 64.8 URL

PointPillars

Model FPS Validation MAP Validation NDS Link
centerpoint_pillar 31 50.3 60.2 URL

nuScenes 3D Tracking

Model Tracking time Total time Validation AMOTA ↑ Validation AMOTP ↓ Link
centerpoint_voxel_1024 1ms 64ms 63.7* 0.606 URL

*The numbers are from the centerpoint_voxel_1024 config (before the sync bn bug fix + z axis augmentation). Current detection models should perform slightly better.