Your selfie-style video could soon be all it takes to turn your computer or smartphone into a mental health-monitoring device, according to researchers at the University of Rochester who have developed a software program for just that purpose.
Leaders of the research team describe the program as a "quiet observer" of your behavior as you go about your screen time, be it work or surf.
Unlike other mental health monitoring methods, you aren't required to pour out your emotions or describe what you're feeling, nor is any supplementary gear required.
It works by monitoring subtle changes in the skin tone on your forehead, which gives it the information it needs to measure your heart rate.
Your heart rate, your blinking frequency, the degree to which your pupils are dilated and your head movements are examples of what the program uses to assess your state of mind.
In a study, participants' Twitter behavior was analyzed, taking into account not only what they tweeted and the tweets they chose to read but also their keystroke speed, how fast they scrolled and the frequency at which they clicked the mouse.
The content of their tweets, however, is treated with more importance than the other factors, say the researchers, whose paper was presented last week at the American Association for Artificial Intelligence conference in Austin, Texas.
To calibrate the system and generate a measurable reaction, UR computer science professor Jiebo Luo jabbed at the emotions of 27 participants by sending them messages and tweets intended to ignite them one way or the other.
Using subjects' reactions to gauge the emotions spurred by positive or negative content in tweets, they ran the data through the program to test whether it could actually analyze emotions.
Currently, the program categorizes emotions as being positive, negative or neutral, although Luo says he hopes to refine the program's analytical capabilities to discern the difference between emotions such as sad or angry -- both of which would currently be classified as "negative" by the program.
