Repository navigation
Anomalib image_threshold with recall=1 #2642
Unanswered
mohamed-s7
asked this question in
Q&A
Replies: 2 comments 3 replies
|
@mohamed-s7 were you able to achieve your objective. I am also getting the image_sensitivity as 0.6 where there is 100% recall but poor precision with a lot of false positives. I am using Patchcore on custom dataset. Any suggestion on how to find the optimal threshold with 100% recall and some False positives are fine? |
3 replies
|
@mohamed-s7 I am using wide_resnet_50_2 with appropriate preprocessing that bypass the default 256x256 resizing with 96x96 resizing. I will try to analyze the FP and their reasons and see how it goes. and thank you for your cooperation |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
in anomalib PaDiM, model.image_threshold = 88.8 for the best F1 score.
but I want the recall = 1 even if I lose a little bit in precision, so I want recall = 1 and maximise the precision.
how can I find that threshold?
using just pred_scores that are between 0 and 1, and labels the threshold that I am getting is 0.4 (value between 0 and 1) but manually I found the threshold that makes the recall = 1 and maximaizes the precision is 77.8
All reactions