search

TuSimple / centerformer

Implementation for CenterFormer: Center-based Transformer for 3D Object Detection (ECCV 2022)
MIT License
292 stars 28 forks source link

Some details to discuss #10

Open Liaoqing-up opened 1 year ago

Liaoqing-up commented 1 year ago

Thank you for open-sourcing your work. I was wondering, why you use x_up(the current frame's bev feature) other than x_up_fuse(the sequential frames through spatial-aware fusion) as center query embedding ? Apologies if I missed it in the paper.

edwardzhou130 commented 1 year ago

Sorry for the late reply. There is no empirical reason for me to choose x_up rather than x_up_fuse. Center classification and box regression need two different types of information from previous frames, so I want to avoid mixing them.