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
Is your proposed algorithm trained in the supervised way? I see that the input of STTR contains GT disparity and occlusion region during training phase.
The text was updated successfully, but these errors were encountered:
OK, I get it. The GT disparity of Relative Response Loss paper is obtained by SFM automatically, so they call it self-supervised training scheme. But yours is provided by the dataset itself, so it's supervised. Am I right?
Is your proposed algorithm trained in the supervised way? I see that the input of STTR contains GT disparity and occlusion region during training phase.
The text was updated successfully, but these errors were encountered: