Thursday 26 November 2015

The Computer That Knows what you are Thinking


Emotion Reading Computers

We as humans have the advantage of masking our innermost feelings whenever the need arises. However, there could be a change now since the computers have been enhanced with not only at recognising faces but also in comprehending what could an individual be thinking in fact. Daniel McDuff has been working ona project where he has been creating emotion reading computers, at the Massachusetts Institute.

 This could ultimately lead to machines which may have emotional intelligence or even everyday objects that have the potential to empathise with our moods, like a mirror that knows how one may be feeling about the way you look, a fridge which could offer food that may match one’s state of mind or a car which may recognises when they are stressed.

The method adopted by Dr McDuff has been progressing through a basic webcam which tends to detect a range of different facial movement from frowning to smiling. It then translates into seven of the most normally recognised emotional states namely, sadness, amusement, surprise, fear, joy, disgust and contempt.

The computer learns from a large database of four million video from volunteers as well as paid-for market researchers, in different emotional states and the procedures are regularly updated as well as tested against real world situations

Integrate Voice Analysis& Other Means of Physical Well being

The next step is to integrate the voice analysis and other means of physical wellbeing like the heart rate and the hand gestures. So far the data has revealed that there are huge differences in emotional responses among men and women and in the various age groups and demographics.

Dr McDuff informed BBC that there are significant differences in various countries in the way people tend to express themselves. He further added that `in collectivist cultures where the family group seems more important, people are expressive in small groups while in more individualistic cultures namely Western Europe, the stress is on building relationships with strangers and people get more positive around people who they are less familiar with’.

Besides this, he also observed that when people tend to mask their real feelings, the computer is capable in recognizing subtle difference. He states that if they are frustrated, they will often smile, however, that smile would be different from the smile when one is genuinely amused.

For Online Education

Another field which the system could be utilised is for the purpose of online education where it could provide advice to tutors on how students seem to be managing and understanding the work. The system which Dr McDuff together with his team has been researching had been tested with the BBC’s audience measurement group offering perceptions to the response to different shows.

 He revealed that among those tested was a prime time comedy show which went down differently with the different demographics, ethnicities and ages. He now intends to use the technology for mental health in partnership with Affectiva, a MIT by-product for which he is the research director.

He states that it could be scary to imagine that someone could measure one’s emotions without the person realising it and hence it could be important to think about the social impact of such technology. It is essential that everyone actively decides to share their data.

 Affective computing seems to be a growing area and companies like Creative Virtual are creating tools for customer service technology which enables a computer to tell the difference between a customer who seems to be upset and the one who is not.

No comments:

Post a Comment

Note: only a member of this blog may post a comment.