Computer Better than Humans at Recognizing Fake Emotions

May 4, 2014

Unlocking Word Meanings
 今日の単語・フレーズ

Read the following words/expressions found in today’s article. 

1. genuine /ˈdʒɛnhttp://static.sfdict.com/dictstatic/dictionary/graphics/luna/thinsp.pngyuhttp://static.sfdict.com/dictstatic/dictionary/graphics/luna/thinsp.pngɪn/ (adj.) – real; not pretending
Example: I felt her genuine pain as she cried for a lost loved one.

2. immerse /ɪˈmɜrs/ (v.) – to put something into a liquid
ExampleI felt better after I immersed myself in the hot spring for half an hour.

3. lukewarm /ˈlukˈwɔrm/ (adj.) – not very warm
Example: The hot coffee had gone lukewarm after being left on the table for 20 minutes.

4. frigid /ˈfrɪdʒhttp://static.sfdict.com/dictstatic/dictionary/graphics/luna/thinsp.pngɪd/ (adj.) – extremely cold
ExampleSometimes, it can be frigid in deep winter.

5. spot /spɒt/ (v.) – to recognize, distinguish, or detect
Example: The child cannot spot the difference between a square and a rectangle.


Article
 ニュース記事

Read the text below.
Scientists have developed a computer system that can recognize genuine expressions of pain better than humans.

A study published in Current Biology highlights a computer program that can spot distinctive aspects of facial expressions that people often miss.

The scientists asked 25 volunteers to record two videos. In the first video, the volunteers immersed an arm in lukewarm water and pretended to be in pain. In the second, the participants immersed an arm in frigid ice water. This time, they experienced real pain, but they were not instructed on what to do with their facial expressions.

The researchers then asked another set of 170 volunteers to identify who among the first set of participants were in real pain and who were not. Based on results, the second set of participants was only 50 percent accurate in their judgment. Because of this result, the researchers trained the 170 volunteers how to tell when someone was faking pain. Even after the training, results showed that they can only manage 55 percent accuracy.

On the other hand, the newly-developed computer system registered an 85 percent accuracy rate. This device has been programmed to recognize combinations of facial movements. It can tell if an expression is real or fake. It uses a camera that can take videos and capture photos.

The study says that the computer can spot distinctive aspects of facial movements, particularly mouth movements, better than humans do. According to one of the researchers, the computer can assess what 20 facial muscles are doing in every frame of the video.

The researchers believe that this system can help recognize the emotional state of patients who cannot communicate very well. It can also be used in law enforcement and in screening job applicants.

Viewpoint Discussion
 ディスカッションテーマ

Enjoy a discussion with your tutor.  

Discussion A

·         Do you agree that computers can be better than humans at identifying real emotions from fake ones? Please explain.
·         In what other ways do you think this invention can be beneficial?

Discussion B

·         Aside from facial expressions, how else can you tell if someone is showing genuine emotion?
·         When do you think it is necessary to fake emotions? Explain.


この記事が気に入りましたか?

May 4, 2014