request for code release

xinshuoweng / SPF2

Official PyTorch Implementation for "Sequential Forecasting of 100,000 Points", In Submission

http://www.xinshuoweng.com/

25 stars 2 forks source link

request for code release #3

Open vivonicole opened 2 years ago

vivonicole commented 2 years ago

Dear Xinshuo,

We would like to treat SPF as a baseline in our work and compare our method with SPF.

We tried to implement SPF by ourselves but the results show it does not make any useful predictions.

Could you please release the code? Thank you so much!

xinshuoweng commented 2 years ago

Hello!

Sorry for the delay! Yes, I am preparing a big code release, including merging this one with one of our ECCV submissions that will be released together in a month! Also, please contact me via email since I might lose track of conversations here. :)

vivonicole commented 2 years ago

Thank you!

benemer commented 2 years ago

Hey,

I am happy to hear that the code will finally be released soon. @vivonicole, if you search for additional baselines for comparison, you can also check out our point cloud prediction CoRL2021 paper with code including a pre-trained model.

Best Benedikt

vivonicole commented 2 years ago

Thank you! @benemer.

benemer commented 2 years ago

Hello!

Sorry for the delay! Yes, I am preparing a big code release, including merging this one with one of our ECCV submissions that will be released together in a month! Also, please contact me via email since I might lose track of conversations here. :)

Hey @xinshuoweng,

any updates on the code releases of SPF2 and S2Net?

In your ECCV submission, you write Note that the original numbers reported in [28] are different due to various changes including improved network architectures, balanced data split, improved implementation of the metrics, etc.. Will you upload the original code of your CoRL 2020 submission as well? This repository is now empty for 2.5 years. It would be nice to have the original implementation to see which changes have been made to improve it.

Thank you very much Benedikt

kaustabpal commented 1 year ago

Hello!

Sorry for the delay! Yes, I am preparing a big code release, including merging this one with one of our ECCV submissions that will be released together in a month! Also, please contact me via email since I might lose track of conversations here. :)

Hi @xinshuoweng, is there any update on the code release?

kaustabpal commented 1 year ago

Dear Xinshuo,

We would like to treat SPF as a baseline in our work and compare our method with SPF.

We tried to implement SPF by ourselves but the results show it does not make any useful predictions.

Could you please release the code? Thank you so much!

@vivonicole Were you able to solve your issue and get a good prediction out of this. Me and my team also did an implementation but we are not getting any useful prediction at all. I am attaching our output. Would love to discuss our architectures. Please send me an email if you are interested: kaustab21_AT_gmail_DOT_com

_AT_ = @
_DOT_ = .