Date of Award

Summer 8-2014

Degree Type

Honors College Thesis



First Advisor

Scott Neal

Advisor Department



Computers through both desktop and mobile devices are only becoming more important in our lives leading us to have more involved and longer interactions with them. Because of this our brains actually classify our involvement with them in a manner similar to our interactions with our fellow humans. This can lead to frustration and anxiety when our computers interrupt our work or pleasure with contextually inappropriate messages, much the same way it would if a friend or co-worker was pushy or rude. A way to solve this issue is to give our machines emotional intelligence, or the ability to recognize and be aware of our emotions. While monitoring physiological symptoms such as skin conductivity and muscle tension is one of the most accurate ways of detecting emotions, it can also be done in a more physically and socially comfortable manner by way of visual and auditory clues. This thesis will create a bimodal system where input is visual information via a still image and auditory information via a clip of human speech. The system will use two existing programs to identify the emotion found in each and, by using a weighted system, return the singular emotion felt.