yuhuayc / da-faster-rcnn

An implementation of our CVPR 2018 work 'Domain Adaptive Faster R-CNN for Object Detection in the Wild'

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

How do you select the ROIS for unlabeled images from target domain ?

alexrame opened this issue · comments

The problem is that, by definition, you do not have groud truth for unlabelled images.

However you need groudtruth

  • in the definition of your blob objects, in lib.roi_data_da_layer.minibatch
  • in order to select the ROIS

I do not see in the code how you made it possible... and I am trying to reproduce the paper in TF

Hi, thanks for the message. We simply use all ROIS from rpn for the domain alignment, thus there is no selection of ROIS invoved.