Intelligent Smartwatch App to Detect Human Emotions
MIT
researchers develop artificial intelligence system to aid in social interaction
using variety of signals
Massachusetts: The researchers from Massachusetts
Institute of Technology(MIT), a privately-owned research university in
Cambridge, Massachusetts, developed a new artificial intelligent software for
the smartwatches to detect emotions of the speaker by analysing their speaking tone.
The system uses speech patterns and vitals of a person to reveal whether
speaker is happy, sad or neutral.
The system uses an algorithm
for real-time analysis of speech and tone and also analyzes blood pressure,
heart rate and body temperature of the smartwatch wearer. The system also reveals
the emotional state in which the wearer was during multiple stages of a
conversation.
The feature demoed on
research friendly wearable Samsung Simband, and by using deep learning and AI
to analyse audio, text transcriptions, and physiological signals, the system
operates with an accuracy of 83%. The researchers at MIT say that with time,
the system will be able to understand more complex and sophisticated emotions.
TechSci Research depicts that
leading organization around the world are spending more on Research &
Development of smartwatches for developing the smartwatches that are more
user-friendly. However, the growing deployment and awareness of the people
towards the benefits of smartwatches is projected to drive their growth, globally,
during the forecast period.
According to recently
published report by TechSci Research,
“Global
Smartwatch Market By Type, By Application, By
Operating System, Competition Forecast and Opportunities, 2012-2021”, the global market for
smartwatches is projected to grow at a CAGR over 25% during 2016-2021. Personal
assistance application dominated global smartwatch market, on account of
increasing usage of third party application in connected devices and rising
demand for technologically advanced products.