Watch all the Transform 2020 sessions on-demand here.
Popular “adult entertainment” site Pornhub will begin using a new AI-powered system that taps facial recognition technology to automatically identify porn stars.
The company has revealed that it will begin teaching the system by feeding thousands of videos into its computer vision model alongside official photos of porn stars. Pornhub will tag vidoes based on the results of this effort, with users requested to “upvote” or “downvote” the porn star tags attached to the video.
Above: Pornhub computer vision model
The new system is designed to complement the existing setup, whereby viewers manually tag videos.
“Ultimately, we want to provide our fans with whatever it is they have a penchant for, and our new model will be able to surface more accurate results for them, which will hopefully keep them coming back for more,” explained Pornhub VP Corey Price. “The easier it is to navigate and access what they want, the happier they will be.”
June 5th: The AI Audit in NYC
Join us next week in NYC to engage with top executive leaders, delving into strategies for auditing AI models to ensure fairness, optimal performance, and ethical compliance across diverse organizations. Secure your attendance for this exclusive invite-only event.
So does this mean that the system could be abused to identify actors who would rather remain anonymous? Well, ignoring the fact that many porn stars use stage names anyway, Pornhub is adamant that its facial recognition efforts will only identify people who are already on its database — all of whom have previously consented.
But Pornhub isn’t stopping with facial recognition. The company said that over the next year it plans to scan all five million of its videos to help with various categorization efforts, such as whether a video was shot outdoors (if so, it will be added to the Public category) or if there are blonde people in a video, for example. It may also be able to detect specific sexual positions.
As AI is being applied to other industries, Pornhub hopes it can streamline its operations and automate many of the processes that have hitherto relied purely on manual labor.
Pornhub isn’t the first porn-related service to dabble in the computer vision arts. Earlier this year, a reverse image-search web app emerged called Pornstar.id, which identifies actors from a pasted URL or an uploaded image. And last year, Megacams introduced a service that lets users upload a photo of anyone to be matched with a lookalike on its live webcam platform.
Facial recognition technology is now being deployed across a broad range of use cases, including public CCTV cameras, so it’s not surprising that the porn industry is taking note.
“Truth be told, we’re growing at an exponential pace — we currently receive 80 million visitors each day, and that number will only continue to climb,” continued Price. “We had an antiquated system for tagging videos that just wasn’t cutting it anymore. Now, with the influx of technology, particularly regarding AI and computer vision, we’re able to update our system and implement a model that will significantly improve content tagging.”
