This repository has been archived by the owner on Jan 13, 2022. It is now read-only.
Use size and compression as metrics #575
Labels
✨ goal: improvement
Improvement to an existing feature
🙅 status: discontinued
Not suitable for work as repo is in maintenance
🏷 status: label work required
Needs proper labelling before it can be worked on
Problem
Images with low resolution or high compression sometimes show up in the first page of results, even with popularity boosting.
This issue blocks on consuming outbound data from the web crawler.
Description
We should heavily weigh down results with low resolution and high compression. Both of these metrics can be distilled into a single "quality_penalty" value (high compression OR low resolution will result in higher quality penalties). The thinking here is that small resolution or high compression are strong indicators that an image is not worth showing, but high resolution and low compressibility do not necessarily correlate with relevance.
The text was updated successfully, but these errors were encountered: