Measure the target's detail at the size candidates are rendered at - #95
Merged
Conversation
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The detail objective read the reference at the pixel comparison's scoring resolution (256 long side) and candidates at their render resolution (700), and compressed size grows with pixel count. The same duck image reads 7,607 at 256 and 41,602 at 700 — 5.5x.
So every candidate was charged for the resolution gap rather than for busyness. In the run this was caught on,
detailwas non-zero on 2410 of 2410 candidates, median 0.943: every candidate read as roughly twice as busy as a target it actually matched. The objective was measuring the wrong thing entirely, and would have pushed the search to strip detail it should have kept.The reference now comes from
original_png_bytes, which is the same size candidates are rasterized at. The resume path already rendered at that size, so it needs no change.A test pins the property, since the comment asserting it was already there and I wrote the bug under it anyway.