Question about downstream

xingyizhou / CenterNet

Object detection, 3D detection, and pose estimation using center point detection:

MIT License

7.29k stars 1.93k forks source link

Question about downstream #1042

Open Younger330 opened 6 months ago

Younger330 commented 6 months ago

Hello,

I noticed that the loss calculation is performed on downsampled images within the network. Is it necessary to compute the loss on downsampled images, or could it also be effectively done on the original resolution images?

Looking forward to your response.

Thank you!