So this is very off-topic, isn’t really anything covered by the current objectives of the MetaBrainz project but us data nerds here might know of something already.
We can currently fingerprint audio files for easy identification which can help with a wide variety of tasks including finding where an item of audio originated from; and of course for the purposes of our objectives it only works with audio content.
What I wonder is if there are other similar fingerprinting technologies, and in turn databases to support them, for other mediums.
I know Google has reverse image search, Youtube has some kind of fancy closed-source scene detection function and then there are services like TinEye which charge for using their tools (although you can do basic queries on the website).
Is there anything in the world of open source that can do these things, any existing projects anyone knows about?
4 Likes
Interesting question. Could I ask for the intent? There are alternatives to Google Lens and TinEye, but as far as I know, the pickings are few and not as productive or accurate.
TinEye is free for personal use, although it is proprietary code. I would think that both TinEye and Google might be open to extending their free product to a non profit on some level, although the non open source issue would for sure be limiting.
Honestly, it comes from a point of being a data nerd; I really like the idea of having systems which can identify media accurately and that are then link to databases.
What can be done with that information is up for discussion; it could be used to enforce image copyright, or even in a lighthanded way prove an images origin. It could be linked to other images that are similar, in a way similar to revisions.
1 Like
I am very interested in the input of others on this topic. I am also doing some research into FOSS for this. Cheers for opening this thread.
2 Likes
For those concerned, during a short tea-break I had a look on ye olde Github and came across this:
Further R&D required later on 
1 Like