Custom AI Face Recognition

written by

  • Derek Ohanesian

published

We have enhanced our AI face recognition feature, first launched earlier this year for celebrity facial recognition, to enable users to customize who the AI recognizes and search for any face in their library. Creative ops users often spend hundreds of hours each month manually tagging images and organizing photoshoots by talent. Now, with custom AI face recognition, you can automate this process. Tag a face once, and Globaledit’s AI will tag that face correctly across your entire library.

How Does Custom AI Face Recognition Work?

  1. Upload training images: Start by uploading a set of images featuring the individuals you want to recognize. These images will serve as the training data for our AI model.
  2. Tag unknown faces: Simply tag a few of the faces in your training images with the appropriate names, and our AI will learn to recognize these individuals.
  3. Automatic tagging: Once your model is trained, it will automatically tag newly uploaded photos and tag recognized faces with the correct names. You can also go back and rescan your previously uploaded assets.

Enabling New Workflows

Managing large photo libraries can be a daunting task, especially when it comes to organizing and tagging images with the correct metadata. Custom AI Face Recognition will be a game changer for anyone working with large libraries of talent, and we see it being especially impactful to industries like:

  • Entertainment: Globaledit is used extensively in the entertainment industry to manage millions of photos and videos for TV and movies. Now those assets can be automatically categorized by the talent in the image.
  • E-commerce: E-commerce brands who work with the same talent can easily look up photos by model name, enabling more efficient workflows.
  • Live events: Live production teams don’t have time to organize their metadata on the fly and now they don’t have to – images can be categorized automatically on upload to Globaledit, allowing live teams to quickly find and publish photos of the right people.

Metadata Schema Customization

Globaledit supports the workflows of customers in a variety of industries, and face recognition is no different. AI Name Tags may go into an industry-standard metadata schema for one customer and a custom schema for another. Our APIs allow us to build custom metadata integrations for face tagging for each customer, increasing interoperability with other platforms and enabling new workflows.

Custom AI Face Recognition allows you to train an AI model to recognize faces of talent and tag those faces with the correct name automatically. We are already seeing the feature shave hours of work off of organization workflows and are already looking ahead to see how we can evolve the feature even further.