Researchers from the Computer Science and Artificial Intelligence Laboratory (CSAIL) at the Massachusetts Institute of Technology (MIT) have developed a device that they say can read people’s emotions without the need for them to say or do anything. Such a capability opens up a range of possible applications, and the inventors are already envisioning its use in “smart homes” that can adjust atmospheric conditions in response to how people are feeling.
The device, known as EQ-Radio, emits wireless signals that reflect off people’s bodies before being detected by sensors in order to record information about their heart rate and breathing. By breaking this data down using “beat-extraction algorithms”, the machine is able to analyze the small variations in the intervals between individual heartbeats.
The researchers say that this enables EQ-Radio to predict whether a person is happy, sad, angry, or excited with an accuracy rate of 87 percent. In a new study, they explain how they developed their algorithms by training EQ-Radio on volunteers as they watched video clips or listened to music designed to evoke particular emotions.
If brought to market, the device could potentially be used by movie studios or advertising agencies to gather data about the emotional responses of audiences. Alternatively, smart houses could use this data to control lighting and temperature according to how people are feeling.
Co-developer Dina Katabi also envisages potential healthcare applications for EQ-Radio, explaining that “our results could pave the way for future technologies that could help monitor and diagnose conditions like depression and anxiety.” Similarly, Katabi’s colleague Fadel Adib says that the device’s ability to measure individual heartbeats could improve the standard of health monitoring, claiming that “by recovering measurements of the heart valves actually opening and closing at a millisecond time-scale, this system can literally detect if someone’s heart skips a beat.”
1
