search

lvwj19 / PPR-Net-plus

PPR-Net++: Accurate 6D Pose Estimation in Stacked Scenarios
Apache License 2.0
36 stars 6 forks source link

Pretrain model for testing #5

Closed rcao-hk closed 2 years ago

rcao-hk commented 2 years ago

Hello,

Thank you for your excellent work! Could you provide pre-trained models for Sileane and IPA datasets?

Best, Rui

lvwj19 commented 2 years ago

We have provided a pretrained model (but not the same as the models in our papers, which were trained in tensorflow) for bunny object of IPA datasets in PPR-Net-plus/h5_dataset/bunny/bunny_pprplus.tar. As for other objects of Sileane and IPA datasets, you can train the models by yourself.

rcao-hk commented 2 years ago

Hi, Thank you for your reply. Could you provide the pprnet and pprnet++'s pose estimation results for the Sileane dataset and IPA real datasets ( Ring Screw and Gear Shaft) for evaluation and comparison? Thank you.

Best, Rui

lvwj19 commented 2 years ago

You can find the pose estimation results in PPR-Net++: Accurate 6D Pose Estimation in Stacked Scenarios.