Closed laserwave closed 2 years ago
Hi,
I am confused how we can do inference on real data with the released pretrain models.
In the eval or inference function in view_select_a2j.py, it requires the 3d joints to do depth image cropping.
I can't understand this. How to get the 3d joints output with only depth image input?
@laserwave , Did you solve the issue? If so, can you tell me the way of doing inference on real data please?
Hi,
I am confused how we can do inference on real data with the released pretrain models.
In the eval or inference function in view_select_a2j.py, it requires the 3d joints to do depth image cropping.
I can't understand this. How to get the 3d joints output with only depth image input?