Sales of fake goods are an enduring nuisance on Facebook. Everyone has seen spam posts in groups and on pages, selling dodgy-looking sunglasses, or designer brands for suspiciously cheap prices.
But the social network may finally have a way to stamp them out.
In a patent application published last Thursday, Facebook outlines technology that is capable of recognising the content of images and flagging them if necessary.
This could, it suggests, be used to spot “a logo that an advertiser may have incorporated into its advertisements without permission.”
This means the likes of Ray-Ban would be able to track down rogue sellers using its logo to flog their wares — even if they have carefully avoided using its name in the post to avoid text filters.
It’s worth noting that this is just a patent right now, and tech companies file thousands of patents every year that never make their way into finished products. But it does show that Facebook is at least considering how to tackle the problem.
Sometimes, the social network says, “images include proprietary content such as items subject to trademark, copyright, or patent protection and for which the content poster does not have the appropriate rights to use the content. In addition, these images may contain other information which can be valuable to classify such as who appears in the images, what letters are being shown, whether the content includes objectionable portions, etc. However, due to the sheer volume of images, it is infeasible for each image to be analysed by a human.”
The tech could also be used to track down the unauthorised sharing of copyrighted images. The company has a well-documented free-booting problem: Rogue pages stealing popular videos and uploading them to grow their following, frustrating original content creators.
These Facebook pages often also steal and recycle viral images — but measures like this could finally help put a stop to it.