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

reproduce the nuscenesces tensorboard result #11

Open
TianhaoFu opened this issue May 11, 2022 · 4 comments
Open

reproduce the nuscenesces tensorboard result #11

TianhaoFu opened this issue May 11, 2022 · 4 comments

Comments

@TianhaoFu
Copy link

TianhaoFu commented May 11, 2022

hi,
I wonder to know how to reproduce the vista nuscenes leaderboard result?

thanks!
@AndlollipopDE @lzhnb @szubing @Yuxin-Wen

@TianhaoFu
Copy link
Author

also i test the provided model and got such performance, which is a litter lower
mAP: 0.6283 mATE: 0.2925 mASE: 0.2452 mAOE: 0.2573 mAVE: 0.1980 mAAE: 0.1967 NDS: 0.6952

could you help me ,thanks very much! :)

@AndlollipopDE
Copy link
Member

Are these performances on val set? Or you submit results to test server? With double flip testing?

@TianhaoFu
Copy link
Author

TianhaoFu commented May 17, 2022

Are these performances on val set? Or you submit results to test server? With double flip testing?

thanks for your reply!

yes, this is performance on val set with tta! @AndlollipopDE

@AndlollipopDE
Copy link
Member

AndlollipopDE commented May 17, 2022

Then these performances are exactly the same as we reported in the README. One thing to know is that the performances on test server are different from the ones on val set. If you upload the test set results to the server, I believe it will produce the same performance as the ones shown on the leaderboard.

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