search

suyz526 / ZebraPose

The implementation of the paper 'ZebraPose: Coarse to Fine Surface Encoding for 6DoF Object Pose Estimation' (CVPR2022)
MIT License
122 stars 20 forks source link

extract features from the ROI #35

Closed hrkita closed 9 months ago

hrkita commented 10 months ago

Hi,

I want to ask, after the Fcos detector obtains the results of Bboxes, how does the ResNet-34 network extract features from the ROI region of interest（e.g. ape）? Is this implemented by the pre-trained model in the pretrained_backbone folder?

Best, Joa

suyz526 commented 9 months ago

Hi Joa,

The dataloader will use the bounding box to extract the ROI, as implemented in #bop_dataset_pytorch.py,

You just need to provide the path of the JSON file, which stores all bounding boxes, to the config file.

Best, Yongzhi

hrkita commented 9 months ago

I have already understood the code now, thank for your enthusiastic respons3 ^_^