Is it possible to use GLAMR by providing a dataset of videos?

NVlabs / GLAMR

[CVPR 2022 Oral] Official PyTorch Implementation of "GLAMR: Global Occlusion-Aware Human Mesh Recovery with Dynamic Cameras”.

https://nvlabs.github.io/GLAMR

Other

359 stars 33 forks source link

Open venturaEffect opened 2 years ago

venturaEffect commented 2 years ago

Hi everyone!

First of all congrats for your amazing work!

My question is simple. Is it possible to GLAMR by providing a dataset of videos?

Another question is. Is there any more improvement achieved on this amazing work?

If it is not possible to add a dataset of videos to this model. Is there any other option?

Thank you so much for your time!