search

SxJyJay / MSMDFusion

[CVPR 2023] MSMDFusion: Fusing LiDAR and Camera at Multiple Scales with Multi-Depth Seeds for 3D Object Detection
Apache License 2.0
167 stars 10 forks source link

TTA #3

Closed Monoclinic closed 1 year ago

Monoclinic commented 1 year ago

Hello, thanks for your great work. I found that your TTA (scaling/double flip) is different from other methods (rotation/double flip) on nuscenes leaderboard, could you please tell the reason?

forwchen commented 1 year ago

For lidar-only models, we found that rotation is useful in TTA. For MSMDFusion (lidar and image), we need to synchronize augmentation for both real 3D points and the virtual points from images, and the virtual points inevitably have some noise. We found that the rotation augmentation brings less improvement than the others (scaling/flip) in this case. So we simply combined scaling and flipping in our final TTA setting.

Monoclinic commented 1 year ago

Got that. Thanks for your reply.