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
Hi,
I was implementing some models to do VQA and found your repo really useful. However, it seems like I can only get the FRCN_FEAT and BBOX_FEAT as image inputs to the model. Is there any way to take the original images as inputs and not the extracted features?
The text was updated successfully, but these errors were encountered:
Hi,
I was implementing some models to do VQA and found your repo really useful. However, it seems like I can only get the FRCN_FEAT and BBOX_FEAT as image inputs to the model. Is there any way to take the original images as inputs and not the extracted features?
The text was updated successfully, but these errors were encountered: