search

LZGMatrix / CDPN_ICCV2019_ZhigangLi

The official training code of ICCV2019 paper "CDPN: Coordinates-based Disentangled Pose Network for Real-time RGB-based 6-DoF Object Pose Estimation".
Apache License 2.0
85 stars 16 forks source link

extend to occlusion datset #7

Open helloyuning opened 1 year ago

helloyuning commented 1 year ago

Dear author,

As title, if I want to use linemod-occlusion dataset to train model, how can I prepare datset for the training stage?

Thanks

XiaoHaoPan commented 1 year ago

Did you fix this issue please, I'm having the same issue.

LZGMatrix commented 1 year ago

@helloyuning @XiaoHaoPan Hi, please refer to issue #6, I will release a short report about how to generate the ground-truth coordinates files in several days, hold on~

XiaoHaoPan commented 1 year ago

Thanks, I wanted to ask a question, if I want to test on the occlusion dataset, is it okay to change the dataset in the config file from lm to lmo?

XiaoHaoPan commented 1 year ago

I tried it doesn't seem to work, can you tell me what I should do?

XiaoHaoPan commented 1 year ago

Please, could you share the OCCILUSION dataset used in your paper？