search

YoungXIAO13 / PoseFromShape

(BMVC 2019) PyTorch implementation of Paper "Pose from Shape: Deep Pose Estimation for Arbitrary 3D Objects"
http://imagine.enpc.fr/~xiaoy/PoseFromShape/
MIT License
170 stars 34 forks source link

Using predict for one image #6

Closed yesimcemek closed 5 years ago

yesimcemek commented 5 years ago

I am trying to use one of your pre-trained models with an object I will use for my project. However, I couldn't find any instructions about how to use the models with another dataset. Could you please explain how to do that? I'm a little lost in your codes right now.

YoungXIAO13 commented 5 years ago

I've added a part called demo, you can see how to apply trained network to new images picturing arbitrary 3D model. Basically, you need to get the right .obj file and generate multiviews from it, then you can save testing images in a directory and run inference to these images.

yesimcemek commented 5 years ago

Thank you for your quick reply. I tried the instructions you provided, but inference.sh gives an error since it couldn't find inference.py. If I'm not mistaken there is no script named inference.py. Could you please check the codes again?

YoungXIAO13 commented 5 years ago

It's added now, sorry for the missing file at the first time : ( You can also train on more images generated from different models in order to get network that is more general. Hope it'll help your project.

yesimcemek commented 5 years ago

Thank you, it works now.