Emotion-Detection Applications Are Built On Outdated Science, Report Warns (eurekalert.org) 18
maiden_taiwan writes: Can computers determine your emotional state from your face? A panel of senior scientists with backgrounds in neuroscience, psychology, computer science, electrical engineering, biology, anthropology, psychiatry, pediatrics, and public affairs spent two years reviewing over 1,000 research papers on the topic. Two years later, they have published the most comprehensive analysis to date and concluded: "It is not possible to confidently infer happiness from a smile, anger
from a scowl, or sadness from a frown, as much of current technology
tries to do when applying what are mistakenly believed to be the
scientific facts.... [How] people communicate anger, disgust, fear, happiness, sadness,
and surprise varies substantially across cultures, situations, and
even across people within a single situation."
Neuroscientist Lisa Feldman Barrett, author of the book How Emotions are Made and behind a popular TED talk on emotion, who was an author on the paper, further elaborates: "People scowl when angry, on average, approximately 25 percent of the time, but they move their faces in other meaningful ways when angry. They might cry, or smile, or widen their eyes and gasp. And they also scowl when not angry, such as when they are concentrating or when they have a stomach ache. Similarly, most smiles don't imply that a person is happy, and most of the time people who are happy do something other than smile."
The American Civil Liberties Union has also commented on the impact of the study. "This paper is significant because an entire industry of automated purported emotion-reading technologies is quickly emerging," writes the ACLU. "As we wrote in our recent paper on 'Robot Surveillance,' the market for emotion recognition software is forecast to reach at least $3.8 billion by 2025. Emotion recognition (aka 'affect recognition' or 'affective computing') is already being incorporated into products for purposes such as marketing, robotics, driver safety, and (as we recently wrote about) audio 'aggression detectors.'"
Neuroscientist Lisa Feldman Barrett, author of the book How Emotions are Made and behind a popular TED talk on emotion, who was an author on the paper, further elaborates: "People scowl when angry, on average, approximately 25 percent of the time, but they move their faces in other meaningful ways when angry. They might cry, or smile, or widen their eyes and gasp. And they also scowl when not angry, such as when they are concentrating or when they have a stomach ache. Similarly, most smiles don't imply that a person is happy, and most of the time people who are happy do something other than smile."
The American Civil Liberties Union has also commented on the impact of the study. "This paper is significant because an entire industry of automated purported emotion-reading technologies is quickly emerging," writes the ACLU. "As we wrote in our recent paper on 'Robot Surveillance,' the market for emotion recognition software is forecast to reach at least $3.8 billion by 2025. Emotion recognition (aka 'affect recognition' or 'affective computing') is already being incorporated into products for purposes such as marketing, robotics, driver safety, and (as we recently wrote about) audio 'aggression detectors.'"
Cons (Score:2)
If this stuff worked, con men (and politicians :) would be out of business.
60 Minutes on china school emotion recognition (Score:4, Interesting)
This weekend on 60 Minutes, one segment showed a chinese classroom with an AI looking at the kids and flagging which ones were "distracted" and which were focused. It looked to me like the kids had already figured out as long as they crossed their arms on their desk in front of them, they could pass as focused.
Re: 60 Minutes on china school emotion recognition (Score:1)
It also leaves out the kids who were smart enough to figure out the material before the lesson completed. It's been known since the art of knowing things was first devised that overachievers often look lazy and slacking.
Reliability (Score:2)
The problem with these sorts of things is that the general public, and often so-called experts, misunderstand the accuracy of the results, and don't really know how to deal with data of unknown reliability. Instead they just think that because it came from a computer than it must be 100% correct.
In reality there will be many false positives. Intuitively it makes sense the computers can detect emotions because we know people detect emotions all the time, some more reliably than others, and clearly there mu