Facial Rekognition has been announced last December as part of AWS artificial intelligence services. Not only is it smart enough to detect faces and objects in an image, it can also search for and compare faces. I can do this through deep learning neural network models, where it goes through billions of images and returns useful attributes per face.
Today, Facial Rekognition has been updated to return a new attribute: estimated age. The service will return an estimate in terms of years. For example, a picture of a person aged 10 may be estimated between 8 to 12 years old. AWS believes this feature can be used for demographics, security purposes, or simply for creating a photo collection of a given time frame.
Facial recognition software has become more affordable and thus widespread in use. Their most common use is security and identity verification for immigration and banking. But with these new features recognizing age and emotion, who knows what new enterprising ways they can be applied?
