search

ZENGYIMING-EAMON / CorrNet3D

CorrNet3D: Unsupervised End-to-end Learning of Dense Correspondence for 3D Point Clouds
GNU General Public License v3.0
137 stars 21 forks source link

About Dataset of Test #1

Closed cainiaoshidai closed 3 years ago

cainiaoshidai commented 3 years ago

Hi,

I download the file of testset.h5, and I found it include ['src_flat', 'tgt_flat', 'label_flat', 'scores_flat', 'sl_002', 'sl_004', 'sl_006', 'sl_008', 'sl_01', 'sl_012', 'sl_014', 'sl_016', 'sl_018', 'sl_02']. I want to know what it all stands for. Thanks.

ZENGYIMING-EAMON commented 3 years ago

Hi, 'sl_xxx' means the soft label for specific error percentage, and the label_flat is the accurate GT label. 'src' and 'tat' means two input point clouds. And scores_flat is not used. If your just want to get the inputs and GT, use 'src_flat', 'tgt_flat', 'label_flat'.

cainiaoshidai commented 3 years ago

Hi, thanks for your reply. I would like to ask you another question. Is the test set you provided the FAUST data set?

Thanks.

ZENGYIMING-EAMON commented 3 years ago

It is the same as the train/test set used in the cited paper 3D-coded.

cainiaoshidai commented 3 years ago

Thanks, I think I know which dataset it is.