Mask_rcnn: How can I compute the AP of my own datasets

Created on 13 Aug 2018  路  1Comment  路  Source: matterport/Mask_RCNN

When I finished training my own dataset, I can't do anything to compute the accuracy as well as the AP of my running.How can I work it out?
Thanks for your attention.

picture Yuekchoi

Most helpful comment

Check train_shapes.ipynb
functions in utils.py including compute_ap will help.

picture MikeXlYang on 13 Aug 2018
👍3

>All comments

Check train_shapes.ipynb
functions in utils.py including compute_ap will help.

MikeXlYang picture MikeXlYang on 13 Aug 2018
👍3
Was this page helpful?
0 / 5 - 0 ratings

Related issues

Jargon4072 picture Jargon4072  路  3Comments
LifeBeyondExpectations picture LifeBeyondExpectations  路  4Comments
techjjun picture techjjun  路  4Comments
wadmes picture wadmes  路  4Comments
JonathanCMitchell picture JonathanCMitchell  路  3Comments