They describe it as ‘passive’ because the subject has no wires or straps attached to them.
Instead they’re filmed answering questions and the video is later analysed using a specially-developed algorithm.
Questions with known answers establish a baseline for anxiety level and emotional reaction….
“What we’re doing is we’re using the video footage. We’re breaking it down into data and then using quant methods, or quantitative math and deep learning from the videos to determine how a persons’ emotional responses are tied to some stimulus so, in this instance, it was questions.”
The technology was first developed for the entertainment industry…
As a way to assess an audience’s reaction to film or TV.
The system analyses facial expressions based on seven basic emotional reactions including anger, happiness and sadness.
“We really want to pan in tight on to her face so we could get a lot of detail in that footage because we are going to be processing it and decomposing the emotional vectors to see the veracity of whether or not she believes in what she is saying.”
The numbers are used to produce an assessment of the seven emotions which update in realtime as the video is played.
“And these would be the points, the maximum anxiety, where we would anticipate that she’s probably being deceitful or being asked a question that she’s not comfortable with the answer.”
If the project is successful it could impact how lie detector tests are done….
And change more than just the movies and TV shows that we all watch.