Short Bytes: A group of MIT researchers has made a wireless device that can detect your emotions without any on-body sensors. Named EQ-Radio, this device performs its functions with the help of RF signal and measuring heartbeats and breathing patterns. Researchers claim that EQ-Radio is 87-percent accurate, outperforming Microsoft’s Emotion API.
Trying to figure out how you feel, even if you try your best to hide it, the researchers at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) has created a device called EQ-Radio. It emits and detects the radio frequency (RF) waves that are bounced off the people to get an idea of their breathing patterns and heartbeat.
The MIT researchers claim that EQ-Radio is 87-percent accurate at detecting whether you are sad, happy, angry, or excited.
This test was performed on 12 participants that sat about 3-4 feet away from the device. To make them feel a certain emotion, the researchers asked them to recall a personal experience. This emotion was established as a ‘ground truth’ that could be used as a baseline for predictions.
It’s interesting to note that EQ-Radio doesn’t need some on-body sensors to predict the results. So, how does the system work?
While one method involves predicting feelings by using a person’s ground truth, the second method uses baselines from the other 11 subjects. Now each subject was asked to recall different experiences that would trigger emotions. The EQ-Radio measured the heartbeats with a surprising accuracy.
In the research paper, researchers note than EQ-Radio is more accurate than Microsoft’s Emotion API.
The researchers believe that it’s an important step in the budding field of emotion recognition. They envision that advancement in such systems will improve as more advanced machine learning methods are inculcated in the process.
For detailed reading, you can find the researcher paper here.
Did you find this article helpful? Don’t forget to drop your feedback in the comments section below.