Human communication can be divided into two subunits: verbal and nonverbal cues. While most sighted individuals are able to communicate in both forms, visually impaired individuals cannot receive most nonverbal messages. The visually impaired are thus handicapped in their social interaction capabilities. To address this handicap, we will develop a computer vision device with facial recognition and expression analysis capabilities in order to relay some nonverbal messages as audio and tactile feedback to a visually impaired user. The device will communicate the names of familiar communication partners as well as their expression states—differentiated as the six universal expressions (i.e. happiness, sadness, disgust, surprise, fear, and anger). The device will be tested with the visually impaired to evaluate its accuracy and usefulness.
How can computer vision be used to improve interaction between the visually impaired and sighted individuals?'
Class of 2012 Gemstone Team FACE:
Questions? Email face.gemstone (at) gmail (dot) com