Unlock Your Media Library with Facial Recognition in Sitecore
Facial recognition is an incredibly valuable and useful functionality to have in your CMS. For marketing teams, it’s revolutionary, not only in that it saves valuable time and money, but also in that it is able to unlock hundreds, if not thousands of images which have been uploaded into a CMS’ media library over the years. Suddenly, it’s possible to search for a person by name and find all the photos of him or her in your database, allowing you to select the perfect photo for a bio page on your site, a newsletter article featuring a particular employee, any number of marketing campaigns and initiatives, and more.
A few years ago, we started exploring how Microsoft Azure’s machine learning (Face and Computer Vision APIs) with Sitecore can make it easy for marketers, editorial teams, and content authors to save time on routine tasks, and integrated the APIs into Sitecore 8.2 and above. With the amount of time that could be cut from routine tasks using machine learning, we determined that teams could instead focus more of their time and efforts on other high-value tasks in their areas of expertise such as audience definition, personalization, and messaging. We created Intelligent Media™, a suite of tools that includes image search, facial recognition, smart crop, and auto alt tagging functionality to help empower digital teams to quickly and easily complete routine tasks. In a previous post, we delved into the image search functionality in depth. Now, we’ll take a look at some of the really powerful things you can do using facial recognition technology.
The Media Library as an Asset: Locate Images of People Through Facial Recognition
Facial recognition technology has made its way into everyday usage through social media. Facebook, for example, has been using it for years to suggest friends to tag in photos. Like Facebook, many businesses are sitting on databases filled with images; unlike Facebook, most businesses haven’t had a way of searching through images of people by name until now.
Typically, most medium to large businesses upload photos of people to their CMS when creating web pages, blog posts, and other digital content. Some organizations have file naming conventions, others do not, and the fact of the matter is, not everyone adds alt tags when uploading images. Ultimately, the CMS ends up with a set of image files over the years that would take a team weeks, if not months, to sift through and manually tag each photo by the people featured in them.
"An organization’s media library is a valuable asset to any content team. The images contained within have typically been curated over years, often housing high-quality images of employees, consultants, and teams."
Whether you have images of doctors at your hospital or business consultants displayed on various pages of your website, the time, effort, and cost it takes to obtain images is not negligible over the long haul. However, most images don’t end up being reusable when they can’t be located in an ever-burgeoning media library.
Now, however, thanks to facial recognition APIs, it’s possible to identify, analyze, and tag faces in photos by training the software. Using the auto alt tag function, you can set alt tags on one to many images using the analysis description.
We’re currently working on smart crop, a tool to create different cropped versions of a person’s photo. With this in-built crop functionality, you’d be able to get headshots from full-body images, for example, without having to manually crop images yourself. The smart crop functionality also works on photos without people.
Use Facial Recognition Software to Enhance Marketing Efforts
Now, let’s examine the practical applications of facial recognition technology using a healthcare organization as an example. Let’s say your marketing team wants to write and publish an article about a specific type of medical treatment, or feature the opening of a new branch of your hospital network, and you want to include specific doctors’ photos. You’ll be able to search your media library for the doctors of your choice and pull up photos that you can use with your article. You’ll also be able to remove images from a site quickly and easily, when a doctor leaves your medical practice, simply by searching your media library for his or her name and finding all pages where their photo is featured.
Additionally, once you train the algorithm, if a doctor appears in a group photo, say from an event, it will show up when you search his or her name. If there are words on a banner with the name of the event, for example, the associated image search functionality would be able to identify the text from the image as well, providing you with information on which event he or she attended. These photos would be invaluable for newsletters and other marketing campaigns and would require minimal effort to identify.
In addition to healthcare, there are a number of other industries where this functionality could prove invaluable including consulting, professional services, publishing, and media.
Next Steps: Add Facial Recognition Functionality to Your Sitecore CMS
Can you envision using facial recognition to improve the efficiency and outputs of your digital marketing and customer experience solutions? Check out our demo below:
We’ll be happy to show you what we’ve done so far and how we can build custom integrations to suit your needs. Reach out to us at [email protected] to learn more or comment below to join the discussion.