The most popular and comprehensive Open Source ECM platform
Let us know more
Artificial Intelligence: Improving the Emotional Intelligence of Algorithms
In order for computers and AI to understand what humans expect of them, it’s important that they can read human emotions. Researchers are trying to do just that. The new field is being called “emotion AI” or sometimes “affective computing“. The field is a mix of computer science, AI algorithms, psychology, and cognitive science. The more AI can understand humans, the better that they will be able to respond to verbal and nonverbal signals.
AI algorithms trained with Emotion AI are able to better interact with humans, begin to understand how emotion affects human communication, understand human-human interactions, and better parse out the meaning of literal and non-literal statements.
Javier Hernandez, a research scientist at the MIT Media Lab, said “think of the way you interact with other human beings; you look at their faces, you look at their body, and you change your interaction accordingly. How can [a machine] effectively communicate information if it doesn’t know your emotional state, if it doesn’t know how you’re feeling, it doesn’t know how you’re going to respond to specific content?”
A branch of Emotion AI is sentiment analysis, the ability to determine the type of emotion a person experienced when they wrote something, typically an on-line comment or short message, like those found on Twitter. But some researchers are trying to determine a person’s emotion by reading their face from images or video or interpreting the meaning of nonverbal interactions of people from video. What humans can immediate sense and pick up about an interaction or dialog can be difficult for an algorithm to understand.
Consider the application of trying to understand the interactions of humans with other humans when driving in automobiles. Stella Batalama, professor at Florida Atlantic University, said that “one of the major issues with the technology of fully or semi-autonomous vehicles is that they may not be able to accurately predict the behavior of other self-driving and human-driving vehicles. This predication is essential to properly navigate autonomous vehicles on roads. The technology tries to teach algorithms the patterns of behaviors of people riding in vehicles.”
