Research suggests that wearable aids can be used to alleviate adverse effects of impaired recognition of facial expressions. In particular, The Austim Glass Project at Stanford is using Google Glass to help children with Autism Spectrum Disorder. Researchers use the Glass in combination with an Android app to help train autistic patients to identify emotions accurately. However, with a starting price of $1500, the Glass is a pricy tool. The goal then was to develop an embedded system with similar functionality, but at a lower price point. In combination with the Picamera, the Raspberry Pi costs less than a $100, making it less risky to insure than the Glass.

What it is

A glasses attachment that takes photos of what the user is seeing, and if a subject's face is detected tells the user what the subject is feeling.

How we built it

Raspberry Pi and Picamera for hardware, AWS SDK and Rekognition API (Python) for processing, Flask for UI

Challenges we ran into

Setting up the Raspberry Pi was initially faulty. Getting Flask to stream photos taken from the Picamera was similarly difficult.

Share this project: