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

pretrained model download link #13

Closed
lucasjinreal opened this issue Nov 18, 2021 · 6 comments
Closed

pretrained model download link #13

lucasjinreal opened this issue Nov 18, 2021 · 6 comments

Comments

@lucasjinreal
Copy link

No description provided.

@gaopengcuhk
Copy link
Owner

We release Container+Single-scale SMCA which can achieve 44.2-44.4 mAP compared with ResNet50+Scale-scale SMCA which achieve 41.0mAP.

@gaopengcuhk
Copy link
Owner

@lucasjinreal
Copy link
Author

@gaopengcuhk thank u. How's the speed compare with r50?

@gaopengcuhk
Copy link
Owner

On Imagenet classification(224), Container is comparable with R50 on FLOPs and speed with much better performance. On Object detection which requires 1333 large resolution, Container is slower than R50. I do not benchmark the speed.

@lucasjinreal
Copy link
Author

@gaopengcuhk why container is slower in large input ?

@gaopengcuhk
Copy link
Owner

All Transformer based models like PVT, Container, SWIN is slower than ResNet 50 on large image.

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