Repository navigation
Add JSON Lines results store and image-level evaluation - #6
Merged
Merged
Conversation
Results are stored one InspectionResult per line, one file per run, and validated on load. Evaluation reports AUROC (Mann-Whitney U, ties as half) and precision/recall/F1 at a threshold chosen on held-out good images for a target false-positive rate, so the test set never sets the threshold. Co-authored-by: Sam Oteng <98087809+samoteng@users.noreply.github.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
vista.store: JSON Lines results store, oneInspectionResultper line, one file per run (as discussed in Inspection Results: JSON files or Postgres for inspection results? #5). Lines are validated on load and errors reportfile:line.vista.evaluation:roc_aucvia the Mann-Whitney U statistic, ties counted as half, no extra dependenciesthreshold_for_fprpicks the threshold on held-out good images to hit a target false-positive rate, so the test set is never used to choose it (ADR 0001)evaluate_imagesreports AUROC plus precision, recall, F1 and FPR at that threshold, with ground truth read from MVTec-style image idsPixel-level metrics (pixel AUROC, PRO) and latency come next.
Testing
pytest(28 passed). AUROC is checked against the pairwise definition, and threshold selection against ties and a zero-FPR target.