CMU’s Facial Recognition Software, Faciometrics

The human face is remarkable. It communicates feelings, identifies a person, and hosts the organs for all five major senses. But could it like posts on Facebook for you?

That is the aim of FacioMetrics, a spinoff of the Carnegie Mellon University Robotics Institute. On November 16th of last year, the company was acquired by Facebook. No acquisition price was disclosed. It was developed by Fernando De la Torre Frade, who had previously worked on IntraFace, the basis for FacioMetrics.

In a statement, De la Torre said “[they’re] taking a big step forward by joining the team at Facebook, where we’ll be able to advance our work at an incredible scale.” He left the university to work at Facebook shortly thereafter.

Ultimately, plans for FacioMetrics and the IntraFace app is to use facial expressions to react to a post on Facebook, as hinted at in a briefing Facebook sent to the website TechCrunch: “[Regarding plans for artificial intelligence research and usage,] future applications of deep learning platform on mobile: Gesture-based controls, recognize facial expressions and perform related actions.”

These features have not been implemented yet, however. For the time

Fernando De la Torre Frade

being, Facebook has used FacioMetrics as a way to improve on their lenses for selfies and live broadcasts. Snapchat, of course, is famous for filters that swap faces or add a dog’s tongue when the user opens their mouth. With this tech at Facebook’s disposal, it may not be long before there’s a whole new breed of flower crowns and anime eyes.

But FacioMetrics could help Facebook usher in a new age of accessibility to their services. According to an article published on the Trib Total Media website, De la Torre brought up using IntraFace to “help monitor the emotional states of patients or detect whether a public speaker was losing the audience’s attention.” Other uses could include spotting drowsy or otherwise distracted drivers, analysis of focus groups for various markets, improved animation and accuracy with video game avatars, and the detection of various mood disorders, such as depression or anxiety.

The IntraFace app

This makes for extremely viable and useful software if put into the right hands- one with the potential to save lives. If IntraFace were to be put into use catching unaware drivers, vehicular accidents would most likely decrease. Or imagine someone with known depression browsing through an app using the tech. IntraFace notices their expression and can alert a loved one or even medical services to help the user. Even if just used to leave reactions on various posts on Facebook, this would allow those with disabilities to use the site just as well as their able-bodied friend.

FacioMetrics is now in use on Facebook with the new camera function. There are still huge potentials for this and other facial recognition software beyond the world of likes and pokes, and the entire technological world will shift along with these developments.

If nothing else, it’ll be nice when my avatar actually resembles me.

Leave a Reply