search

impiga / Plain-DETR

[ICCV2023] DETR Doesn’t Need Multi-Scale or Locality Design
MIT License
192 stars 4 forks source link

Question about the ablation of box relative position bias #7

Open julightzhong10 opened 1 year ago

julightzhong10 commented 1 year ago

Thanks for the great work!

I have a question on the BoxRPB:

In the ablation of paper, when you use center as the encoding points instead of 2-corners, the performance seems very similar to DAB/Condition/SMCA cross attention, have you also try using the 2-corners schema in DAB/Condition/SMCA? Just want to figure out if the major improvement of BoxRPB over DAB/Condition/SMCA is due to using 2-corners instead of 1-center.

Thanks!