You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Thanks for your nice work, I am going to try to use your proposed method in our work, but the current code repository code is based on the implementation of detectron v2, the implementation of the function call is a bit complicated, is there an implementation based on the original DETR code? Or can you explain the key code of refinebox? I think an implementation based on the original DETR might help to increase the impact of this work.
The text was updated successfully, but these errors were encountered:
Thanks for your attention. We don't have an implementation based on the original DETR code.
RefinerMixin describes the interface. The forward_refinement will be called after the detector produces detection results. This method refines the detector outputs (by calling the refine_box method) and returns loss_dict (training) or refined results (inference).
Thanks for your nice work, I am going to try to use your proposed method in our work, but the current code repository code is based on the implementation of detectron v2, the implementation of the function call is a bit complicated, is there an implementation based on the original DETR code? Or can you explain the key code of refinebox? I think an implementation based on the original DETR might help to increase the impact of this work.
The text was updated successfully, but these errors were encountered: