Apple Is Looking At Facial Recognition For Easing Photo-Sharing

Perhaps taking a leaf out of Facebook’s playbook, Apple is looking at ways to automatically share photographs with the people being snapped, according to a new patent filing (spotted by AppleInsider). The driver for such efforts being the sheer volume of snaps smartphone owners are now taking — many of which doubtless languish unlooked at on phone camera rolls.

Much like the Facebook Moments app, which launched this summer and uses facial recognition tech to help distribute photos to the people in them, Apple’s patent — which is entitled ‘Systems and methods for sending digital images‘, and was filed in February 2014 but published this week — describes various methods for streamlining the sharing of photos by linking faces to contact data, also utilizing facial recognition tech.

As AppleInsider notes, Apple already uses facial recognition tech to help users identify and sort their photos, such as in its Photos app for Mac. This patent builds on that by describing ways to link contact data to identified faces, and giving users different options for sharing photos with people in them — such as via email or SMS.

Multiple contact-linking and photo-sharing methods are covered in the patent, including letting users manually associate contact info with faces, and post to social networks as another way to distribute photos.

From the filing:

Facial recognition algorithms may identify the faces of one or more people in a digital image. Multiple types of communication may be available for the different people in the digital image. A user interface may be presented indicating recognized faces along with the available forms of communication for the corresponding person. An indication of the total number of people available to be communicated with using each form of communication may be presented. The user may have the option to choose one or more forms of communication, causing the digital image to be sent to the recipients using the selected forms of communication. An individual may have provided information for facial recognition of the individual to a service. Based on the information, the service may recognize that the individual is in an uploaded picture and send the digital image to the user account of the individual.

Google notably overhauled its own Photos app recently, applying machine learning algorithms to auto-sort users’ camera rolls — allowing people to use natural language queries to search through their shots. Of course the flip-side of letting Google’s algorithm crawl all over your camera roll is it’s sucking up all sorts of personal data based on the intel it can glean from your photos.

By contrast, Apple has been making increasing loud noises about how its business model does not rely on the accumulation of user data — emphasizing this as a differentiator vs ad-driven business models, like Google’s and Facebook’s.

At WWDC earlier this summer Apple debuted an update to its Siri voice assistant, called Proactive, which offers some predictive Google Now-esque features. However the data processing associated with this service is done locally on the user’s device — rather than via cloud processing — drawing a clear line between Apple and the user’s personal data to bolster privacy.

Returning to the facial recognition photo-sharing patent, Apple notes that the associated database linking facial recognition info with individuals’ contact details could be stored locally or in the cloud:

The facial recognition data for the face may be stored in association with the name and address for the face. For example, a record in a database with a facial recognition data field, a name field, and an address field can be created or updated. More or fewer fields may be present. The database may be stored on the client machine 110 or 112, or on the server 118. In some example embodiments, the client machine 110 or 112 accesses the server 118 to retrieve facial recognition data for facial recognition on the client machine 110 or 112. In other example embodiments, the client machine 110 or 112 accesses the server 118 to transmit an image for facial recognition on the server 118.

If Cupertino does implement a facial recognition system for photo distribution, given its strong pro-privacy trajectory, it would be difficult to imagine it being able to implement cloud-based processing or storage of such sensitive user info. Local processing would be more in keeping with its current pro-privacy stance.

That said the patent covers all bases — describing, for instance, a scenario involving synchronizing previously trained facial recognition data (from a desktop picture organization tool, such as presumably Apple’s own Mac Photos tool) via a “central server system” to a mobile device:

Once synchronized, the mobile device can utilize the pre-trained facial recognition data to recognize faces within newly taken pictures. Subsequently, the contact information associated with the pre-trained facial recognition data can be used to share the newly captured photo with known users according to the examples discussed herein.

In another method described, the system could even automatically distribute photos to people identified within them if the user has opted in to auto uploads.

In the push to serve the gods of convenience the cloud looms large, allowing for digital content to be trivially accessed across multiple devices. But so too do privacy implications. So Apple will have some key choices to make as it negotiates how to apply more AI- and machine learning-powered features into its software.

Locking down user privacy when there’s cloud syncing and auto-uploads going on in the background can certainly be a tricky circle to square — as one user of Google Photos recently discovered. So it will be interesting to see how Apple implements a facial recognition photo-distribution feature, if indeed it does decide to turn this patent into a concrete consumer product.