This disclosure describes techniques that leverage the large volume of video captured by smart glasses to improve people identification. Bootstrapping off face recognition, the data collection capabilities of smart glasses or other devices are leveraged to train machine-learning models to utilize other input factors for people identification. Such factors can include, for example, gait, pose, silhouette, etc. The techniques generalize and expand people identification beyond face recognition, and, in doing so, hew closer to how humans recognize people.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.