Training data in YCB-video image

cvlab-epfl / segmentation-driven-pose

Segmentation-driven 6D Object Pose Estimation. CVPR 2019.

GNU General Public License v3.0

184 stars 37 forks source link

Open sandy-ssdut opened 4 years ago

sandy-ssdut commented 4 years ago

Hi，

How many training data be generated from 80K synthetic in YCB dataset？ Did all the 80K synthetic be used， or just a part of them？