Facebook started it. Windows Live Photo Gallery has implemented it since the 2010 version and made it easier with the 2011 version.
What is people-tagging
The feature I am talking about here is the ability to attach a metadata tag that identifies a particular person that appear in a digital image. These implementations typically have the tag applied to a specific area of the photo, usually defining the face or head of the person concerned. It will also become available in current or up-and-coming versions of other image-management programs, photo-sharing services, DLNA media servers and the like.
In the case of DLNA media servers, one of these programs could scan an image library and make a UPnP AV content-directory “tree” based on the people featured in one’s photo library.
Initially the concept, especially the Facebook implementation, was treated with fear and scorn because of privacy invasion. This is because this implementation allows the metadata to be related to particular Facebook Friends and also allows the photo to be commented on by other Facebook Friends. Now the Windows Live Photo Gallery application attaches this metadata in a standardised XML form to the JPEG file like it does with the description tags and geotags. There is the ability to make a copy of this file without the metadata for use in posting to Internet services.
A relevant implementation idea
One key benefit that I would see with this data when implemented with electronic picture frames, HDTVs and similar devices is the ability to overlay the tags over the picture when it is shown. This could be achieved by the user pressing a “display” or similar button on the device or its remote control. Devices with touchscreens, stylus-operated tablet screens or other pointer-driven “absolute” navigation setups could support a function that shows a “people tag” as you touch areas of the image.
Benefit to Alzheimers sufferers
Here, this feature could help people who suffer from Alzheimer’s or other dementia-related illnesses by helping them remember whom their family members or friends are. If the user is using an image-management program or DLNA media-server setup capable of using these tags, they can call up a collection of images of the person they think of and have those images appearing on the screen. If the device has a communications-terminal function like a telephone, one of the images can be used as an index image to remember the correspondent by. This function could be extended by the use of an automatically-updated index image or a screenshow that shows “key” images of the person.
Improving on the idea
To make this work, there needs to be an industry standard that defines how the people-tag metadata is stored on the JPEG file. As well, the standard has to support functions like one or more separate “nickname” fields for each of the people that can be displayed as an option. This is because a person may be known to one or more other people via a nickname or relative-shortcut name (Mummy, Daddy, Nonna, etc).
Another issue is to encourage users to establish consistency whenever they tag up a collection of images. This could be achieved through “batch-tagging” and / or improved facial recognition in image-management tools. This may be an issue if two or more people are tagging images from their own collections to serve a third collection and they know the people via different names.
Once we cut through the hysteria surrounding people-tagging with digital images and focus on using it as part of desktop image-management systems rather than social networks, we can then see it as a tool for helping people remember whom their loved ones are.