Meta has patented an AI system for smart glasses: automatic facial recognition and video clipping.
Meta Platforms Technologies continues to expand the functionality of its wearable devices. The U.S. Patent and Trademark Office has disclosed a new company filing describing a system for smart glasses capable of recognizing faces in real time, analyzing user actions, and automatically creating video highlights of key moments.
The essence of the technology lies in integrating advanced computer vision algorithms directly into the device's optics. The system tracks the owner's gaze, identifies people in the field of view, and assesses their behavior. Based on this data, an AI assistant compiles a selection of the most significant fragments, saving the user from having to manually review hours of recordings.
The patent documentation provides a clear example: a user wearing smart glasses attends a formal dinner, scans the guests, and then receives a ready-made video from the assistant featuring the best moments of the evening. This approach implies not just passive recording, but active analysis of social context and highlighting emotionally or event-important scenes.
This is not Meta's first step toward smart cameras, but this particular filing underscores the company's ambitions in personalized AI. The technology could form the foundation for future generations of Ray-Ban Stories smart glasses and other AR devices, where the key will be not only image capture but also its intelligent processing.
Clearly, Meta aims to shift competition in the wearable electronics market from the realm of "hardware" to the realm of algorithms. However, such functionality inevitably raises privacy concerns: automatic face recognition in public spaces requires clear legal frameworks, otherwise the technology risks facing regulatory restrictions in the EU and several other jurisdictions. Nevertheless, the potential for everyday use is enormous—from creating family archives to professional event filming without an editor.