Researchers from Harvard and the University of Vermont have developed a machine that uses a variety of metrics to accurately assess people's mental health based on their Instagram photos.
The artificial intelligence uses people's photos, meta data and facial recognition tools to identify and track their emotional state. The machine looked at thousands of photos from 166 Instagram users and was found to have a 70% accuracy rate in identifying clinical depression -- more capable than general practitioners. Features of photos including lighting, filter colors and facial expressions are all signifiers that the AI uses to determine who is clinically depressed.
This machine's ability to process a variety of different metrics with relatively accurate outcomes could have potential uses in the field of psychiatry.