Detail about the VFE module

Haiyang-W / DSVT

[CVPR2023] Official Implementation of "DSVT: Dynamic Sparse Voxel Transformer with Rotated Sets"

https://arxiv.org/abs/2301.06051

Apache License 2.0

361 stars 28 forks source link

Closed yipliu closed 1 year ago

yipliu commented 1 year ago

Hi,

Since I'm new to this field, I didn't understand the voxel feature encoding (VFE) module very easily.

It seems that this module (VFE) is a common structure, can you please elaborate on it?

Is it voxelnet or pointpillars?

Haiyang-W commented 1 year ago

Yes, it's a common structure. We adopt the dynamic vfe adopted in Pointpillar with pillar size (0.32, 0.32).