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
I think here mean_y_addition represents the y axis of the added object and mean_y_original represents the y axis of the select object in the original scene, then when the positional addition is a "below" relation (added object -> below -> select object), the mean_y_addition should be greater than mean_y_original. But here the score is given for the opposite. Same problem for the "top" relation.
The text was updated successfully, but these errors were encountered:
zzhang2816
changed the title
Some confusion issues in the code
Some confusing issues in the code
Sep 10, 2024
Thanks for the greate work. But there seems exist some problems in the code:
Function name
editval_code/auto_eval.py
Line 364 in affe95b
It seems should be evaluate_position_replacement, so it can be called in the following code
Import module
editval_code/FID_evaluation/fid_score.py
Line 38 in affe95b
No module named "inception". I change it to "from pytorch_fid.inception import InceptionV3"
For evaluating positional addition
editval_code/auto_eval.py
Line 329 in affe95b
I think here mean_y_addition represents the y axis of the added object and mean_y_original represents the y axis of the select object in the original scene, then when the positional addition is a "below" relation (added object -> below -> select object), the mean_y_addition should be greater than mean_y_original. But here the score is given for the opposite. Same problem for the "top" relation.
The text was updated successfully, but these errors were encountered: