5 ms·
It’s more feasible that an app would analyze the metadata and image itself than the audio, and I don’t think we’ve seen that happening.
by soared 3y ago
It’s more feasible that an app would analyze the metadata and image itself than the audio, and I don’t think we’ve seen that happening.
- lancesells 3y agoMeta analyzes the metadata and image. I would be shocked if they aren't analyzing the audio of Live Photos since their release.
- soared 3y agoFor ad targeting?
- a13o 3y agoYes. Facebook runs all photos thru digital recognition. If you look at the alt text for a photo in the feed it will have a string generated by their photo analyzer. But that's just the front. They also use facial recognition to learn which other FB accounts are in your orbit, or make shadow profiles if they don't recognize a facial fingerprint. That way if that person ever signs up, they match the metadata from the shadow profile. This is all in service of scaling ad targeting surface area, which can be sold by the eyeball to generate massive wealth.
- spondylosaurus 3y agoHow would you turn live photo audio into adtech-compliant metadata? AFAIK there's nothing remotely like that in the OpenRTB spec, which is the standard for all real-time bidding ads.
- saagarjha 3y agoText to speech?
- spondylosaurus 3y agoSo speech gets converted to text, then, but how does that text get turned into usable metadata that follows adtech-specific protocols?
- saagarjha 3y agoWhat does adtech do with all the other metadata they process?
- spondylosaurus 3y agoIf you're actually curious, the OpenRTB spec is open-source. You can see the entire protocol in detail here: https://github.com/InteractiveAdvertisingBureau/openrtb/blob/master/OpenRTB%20v3.0%20FINAL.md https://github.com/InteractiveAdvertisingBureau/openrtb/blob...
- saagarjha 3y agoPerhaps I’m misunderstanding here. This seems like the spec for advertiser bidding. Why is this relevant for how user data is consumed for targeting purposes?