A person dissipates various kind of emotion during the daily conversation. The expression actually depends on the statement of the narrator but, sometimes it is very difficult to perceive someone’s sentiment behind the speech. To unveil the emotion behind someone’s speech, scientists created an artificial intelligence algorithm.

A research team from Massachusetts Institute of Technology’s(MIT) from Computer Science and Artificial Intelligence Laboratory (CSAIL) and Institute for Medical Engineering and Science (IMES) built a wearable app that is programmed with the algorithm. It could be a perfect alternative of polygraph because polygraphs are not really reliable because it has a lot of errors. Their invention was first officially introduced in the official press release of MIT News.

Ph.D. candidate Mohammad Ghassemi, lead researcher of this program said in a statement, “Our work is a step in this direction, suggesting that we may not be that far away from the world where people can have an AI social coach right in their pocket”. Ghassemi and his team equipped a fitness tracker with the app that collects physical and speech data to examine the tonal characteristics.

According to the Verge, this AI algorithm can track emotional changes such as happiness, sadness, anger, fear in five seconds of intervals. During the research, volunteers were asked to wear a ‘Samsung Simband’ with that app installed. Each participant was telling a story wearing the band on their wrist. That app was monitoring the heart rate, physical changes, arm movements and skin temperatures during the story telling.

The result was magnificent, the app worked with 83 percent of accuracy. Graduate student of MIT Tuka Alhanai, who was the part of this research explained that at the end of the conversation, this app will allow anyone rewind to the moment to see when the people around felt most anxious. This product could be used to help people with an anxiety disorder or conditions like autism.

Sadness was detected by long pauses, monotonous vocal tones, increased fidgeting and cardiovascular activity. On the other side, happiness was identified by energetic and varied speech patterns. Researchers are now planning to upgrade their system to perceive more complex emotion patterns.

Related