MichalBusta/E2E-MLT

3 times call to function net.forward_ocr in method process_boxes

AniketGurav opened this issue · 2 comments

Hi, in function process_boxes net.forward_ocr is called 3 times. I am not clear about it.
those lines no are 270,276,381 in train.py

By reading paper, what I understand is the function process_boxes ocr the crops extracted by the
Localization Module LM.
Those crops are extracted from the 1. bounding box coordinate extracted by LM and 2.feature map from one of the layer of LM.

But I am not clear about 3rd ocr call on line 381 above..

I have referred Fig 3 of your paper https://arxiv.org/pdf/1801.09919.pdf for understanding.

Hi Aniket,
3rd call is training on GT boxes. It can speed up training in early stage (since a prediction network does not produce proposals with good overlap)

in short we train on:

  • gt boxes
  • proposals with estimated angle
  • propostals with gt angle (You can see it as extra augumentation)

Hope it helps, Michal

Thank you for reply