graph-rcnn.pytorch
graph-rcnn.pytorch copied to clipboard
Question about position encoding in Relationshipness(nn.Module)
Thank you for a nice work!
I found that there is pos_encoding argument in class Relationshipness (relationshipness.py) I guess it is to regress each relationshipness with considering geometric location of each box. But i found any expression about it in the paper. Did you use pos_encoding to report the experimental result?
Thanks,