An Australian sporting league simulated live crowds using cardboard cut-outs to fill empty stadiums during COVID-19. Fans were allowed to submit their own faces to be used on the cutouts, which resulted in some (mostly harmless) misuse, for example fans submitting their pets' faces

enter image description here

But it also resulted in an instance of nefarious misuse where undesirable faces were uploaded and this went undetected by administrators until after the broadcast.

Although very imperfect, it has been possible for some time to filter text for undesirable strings by way of a lookup against lists of undesirable strings.

By analogy, do any similar open-source lists exist for faces? That is, are there any openly available undesirable face lists that can be used by facial recognition software for the purpose of matching/filtering, or at least reducing the manual (human) workload involved in dealing with (probably) unwanted faces in images by flagging suspect images?

Note: Given any such list(s) would be highly subjective, it would be up to the responsible developer to ensure the appropriateness of lists for any particular use case (just as is the case for filtering obscenities from text). Despite the subjectivity/imperfections, are there any open source lists available for undesirable faces as there are for strings?

  • $\begingroup$ Are you just looking for a Dataset then you may place this question on SE Opendata opendata.stackexchange.com $\endgroup$ – 10xAI Jul 2 at 7:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.