German researchers at the Fraunhofer Institute for Integrated Circuits IIS have developed a real-time facial recognition app for Google Glass that can estimate age, gender, and the basic emotions of happiness, anger, sadness, and surprise. The research firm’s facial recognition software is called Sophisticated High-speed Object Recognition Engine (SHORE) and uses the CPU in Google Glass to process video in real-time.
SHORE relies on a database built on C++ that includes over 10,000 annotated faces. The program uses its library of facial expressions to determine the age, gender, and emotion of the person being viewed through Google Glass. According to company representatives, “The new development from Fraunhofer IIS is the first emotion recognition software in the world to function in real-time with Google Glass.”
Although SHORE’s programming and range of features are impressive, and a precursor to more sophisticated facial recognition apps to come, the impact this will have on personal privacy in public and private settings will no doubt rankle citizens and local government agencies still struggling to reconcile technology that can record personal data on the sly and in real-time. In an anticipation of this dilemma, the team responsible for developing SHORE insists that “none of the images leaves the device.” Names will neither be searched nor populated, only the emotional state, gender and age of the individual.
The research firm goes on to explain that the app will be marketed to individuals suffering from certain types of autism in which the individual cannot identify emotions (Okay, I dig), and also to commercial and market research firms (Okay, I dig not so much).
The app was developed thanks to the company’s participation in the Google Glass “Explorer Program,” and is not yet available to consumers.