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

[Feature] Client Side Hamming Distance Settings for Identify Tasks #5596

Open
kbstash opened this issue Jan 7, 2025 · 0 comments
Open

[Feature] Client Side Hamming Distance Settings for Identify Tasks #5596

kbstash opened this issue Jan 7, 2025 · 0 comments

Comments

@kbstash
Copy link

kbstash commented Jan 7, 2025

Is your feature request related to a problem? Please describe.
False positives frequently occur when using the identify function with various stash-box instances. Partly I believe this is due to some bad hashes on the stash-box side which is understandable. False positives also seem to stem from the phash hamming distance that is set on the server side. In my case I would only like to pull exact matches into my db when using identify. While I understand one of the strengths of perceptual hashing is matching on images with some distance, I believe that allowing the user to decide the distance would be a better use experience.

Describe the solution you'd like
Allow client side hamming distance to be set per stash-box in the identify settings.

Describe alternatives you've considered
Adding a similar setting to the ad hoc tagger settings might also be useful. That interface feels more suited to setting varying distances as the user is validating the matches before saving in most cases.

Additional context
I believe fansdb recently made the server side change to increase the hamming distance from 0. This created well over 1000 false positives in a single identify task. It feels like filling local stash instances with high numbers of false positives can eventually lead back to those bad matches being submitted to stash-box instances if users aren't being diligent.

@kbstash kbstash changed the title [Feature] Short Form Title (50 chars or less.) [Feature] Client Side Hamming Distance Settings for Identify Tasks Jan 7, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant