Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Hi, what's the difference between these two pretrained backbones? #75

Closed
tyxsspa opened this issue Nov 22, 2022 · 2 comments
Closed

Hi, what's the difference between these two pretrained backbones? #75

tyxsspa opened this issue Nov 22, 2022 · 2 comments

Comments

@tyxsspa
Copy link

tyxsspa commented Nov 22, 2022

I found two backbones that used in config, one is v2-99 fcos3d_vovnet_imgbackbone-remapped.pth and another is dd3d_det_final.pth , and dd3d is used for nuScenes TEST, I wonder what's the difference? Thanks~!

@yingfei1016
Copy link
Collaborator

fcos3d_vovnet_imgbackbone-remapped.pth is first pretrained on DDAD15M dataset and further trained on nuScenes train set with FCOS3D.

dd3d_det_final.pth is first pretrained on DDAD15M dataset and further trained on nuScenes trainval set with DD3D.

The 'dd3d_det_final.pth' may have overfitted of the val set, the result will be inflated. So if you report the result on val set,you must use the 'fcos3d_vovnet_imgbackbone-remapped.pth'.

@tyxsspa
Copy link
Author

tyxsspa commented Nov 22, 2022

Got it, thanks!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants