Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Low PQ in inference #119

Open
ankur-chr opened this issue Jun 29, 2022 · 0 comments
Open

Low PQ in inference #119

ankur-chr opened this issue Jun 29, 2022 · 0 comments

Comments

@ankur-chr
Copy link

ankur-chr commented Jun 29, 2022

I am getting the following inference output after 10k training iterations on a ResNet34 pre-trained backbone.
Are the PQ values supposed to be so low? For ex: 0.34 (does this actually mean 34 percent?)

[06/29 20:23:42 segmentation_test]: OrderedDict([('segm', {'AP': 5.71368072249309, 'AP50': 14.698149986496215})])

[06/29 20:25:24 segmentation_test]: {'All': {'pq': 0.3420215735136758, 'sq': 0.6189864762517887, 'rq': 0.4380984066808291, 'n': 19}, 'per_class': {7: {'pq': 0.9477637171822374, 'sq': 0.9526897240282989, 'rq': 0.9948293691830403}, 8: {'pq': 0.5613471605445204, 'sq': 0.7657780667726741, 'rq': 0.7330415754923414}, 11: {'pq': 0.7076619260716286, 'sq': 0.8233854645704008, 'rq': 0.8594539939332659}, 12: {'pq': 0.06882625524072358, 'sq': 0.6423783822467534, 'rq': 0.10714285714285714}, 13: {'pq': 0.11938448060671651, 'sq': 0.6653197617145139, 'rq': 0.17943925233644858}, 17: {'pq': 0.2233735409372342, 'sq': 0.6011939301796417, 'rq': 0.37154989384288745}, 19: {'pq': 0.17808683224783267, 'sq': 0.630980069429821, 'rq': 0.2822384428223844}, 20: {'pq': 0.49446154724610397, 'sq': 0.7018699143795368, 'rq': 0.7044917257683215}, 21: {'pq': 0.743941329201978, 'sq': 0.8195109901818628, 'rq': 0.9077868852459017}, 22: {'pq': 0.21917204129954795, 'sq': 0.711410888152631, 'rq': 0.30808080808080807}, 23: {'pq': 0.8017755512307644, 'sq': 0.8956862895991033, 'rq': 0.895152198421646}, 24: {'pq': 0.3162603266031253, 'sq': 0.7006327371718746, 'rq': 0.45139244831710235}, 25: {'pq': 0.2117730720003218, 'sq': 0.6146195924219866, 'rq': 0.344559585492228}, 26: {'pq': 0.4722979928801463, 'sq': 0.7887466101210944, 'rq': 0.598795591410067}, 27: {'pq': 0.0, 'sq': 0, 'rq': 0.0}, 28: {'pq': 0.2772966560348428, 'sq': 0.8078931421015132, 'rq': 0.3432343234323432}, 31: {'pq': 0.0, 'sq': 0, 'rq': 0.0}, 32: {'pq': 0.0, 'sq': 0, 'rq': 0.0}, 33: {'pq': 0.1549874674321157, 'sq': 0.6386474857122791, 'rq': 0.24268077601410934}}, 'Things': {'pq': 0.179076939368819, 'sq': 0.44381744594109346, 'rq': 0.24758284058323124, 'n': 8}, 'Stuff': {'pq': 0.4605267619826625, 'sq': 0.7463821346595672, 'rq': 0.5766551820245366, 'n': 11}}

@ankur-chr ankur-chr changed the title Sample output? Low PQ in inference Jun 29, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant