dingdingcai/OVE6D-pose

viewpoint codebook issue

lgq-gpu opened this issue · 5 comments

Hello, I'm sorry to bother you, I'd like to know with sex which code can get the viewpoint codebook of obj mesh object? @dingdingcai

Hi, the viewpoint sampling and encoding can be found here

def OVE6D_codebook_generation(model_func, codebook_dir, dataset, config, device):

and

def viewpoint_sampling_and_encoding(model_func, obj_model_file, obj_diameter, config, intrinsic, device):

Thank you for your reply, I'll check it out

Hello, I would like to ask if the input mesh model must be normalized? @dingdingcai

Hello, I would like to ask if the input mesh model must be normalized? @dingdingcai

I didn't try it, but I think mesh normalization is not mandatory and you have to make sure that the rendered depth is consistent with its real observation.

I see. Thank you for your reply. It's very disturbing @dingdingcai