Peek into the mind


CHANDIGARH: In his 1998 media address, former US president Bill Clinton strongly denied allegations of having sexual relations with a female employee of the White House. On the face of it, it seemed Clinton had been falsely implicated, but recently, technology has helped analyze his emotions at the time of his address.

In the more than five-minute video, a billboard was superimposed, which added its own algorithmic software to evaluate Clinton's underlying feelings. It demonstrated Clinton's "strong internal conviction", "desire to convince" and "desperation to win the support of Americans," while addressing the media.

Such analyses has become possible with the help of an emotion detection system that can easily understand the meaning of intonations of voices and represent them. The software is being widely used by various global companies, especially BPOs and consultancy firms.

Many interesting offshoots of this emotion detection software have been developed in the form of various apps. Some companies at IT Park, Chandigarh, are also using the software. Nitin Moudgil, who runs a software development firm at IT Park, said, "My clients are based in France and Sweden. Of late, I have started using an emotion detector software. It is really smart and helps you get a fair idea about the mindset of your potential client. Besides video calls, it also helps in text chatting."

Working, future prospects

A billboard is superimposed on the video or audio clip, that adds its own algorithmic evaluation software to understand original feelings. The software is an emotion detection system that strikes correlations between voice intonations before presenting a conclusion.

The algorithms could rank the enthusiasm of candidates by evaluating his reading or way of speaking and also analyse whether he is an introvert or extrovert inquisitive or more practical.

They can also evaluate a person"s mood and suggest a playlist of songs. The algorithms can help computers understand what humans feel, a field known as "affective computing".

New techniques have been developed in computational voice analysis that help machines to identify with the help of writing.

Helps to detect 400 different moods:

There are emotion detection software that can help detect 400 different mood variations, from happiness to sadness, optimism to dejection and seriousness to casualness.

Research and development is on and efforts are being made to introduce emotional understanding into practically everything humans do; including face-to-face interactions, driving cars or playing games.

Steve Jobs showed signs of loneliness and fatigue in last interview:

In one of his last interviews, while Steve Jobs recollected old memories and talked about his feelings while the i-phone was developed, he was actually feeling lonely and fatigued. This was deduced by an emotion detection software. The nearly two-minute youtube video has gone viral across the world.

"I had this idea of being able to get rid of the keyboard, type on a multi-touch glass display and I asked our folks, could we come up with a multi-touch display that I could type on, I could rest my hands on and type on. It (i-phone) was amazing," said a smiling Jobs in the interview. However, the billboard superimposed over this video said Jobs was in the throes of various emotions at the time. There were "conflicts between urges and self-control, loneliness, fatigue, emotional frustration, sadness mixed with happiness, possibly nostalgia." It is said in a ticker run over Jobs"s head.

Tel-Aviv based start-up behind this software:

Research in the direction of evaluating voice intonations to understand underlying emotions and meanings, was started way back in 2011. However, the first success was made in 2012, when a Tel Aviv (Israel) based start-up offered a nascent technology to study consumers" emotions in real time. Their initial clients were call centres. Gradually, more companies jumped into the field. At present, many softwares and apps are available for this.

Emotional analytics engine:

Most of the emotion detection software have their own algorithms that work on an independent emotional analytics engine. From the parent company, license of these algorithms are given to third parties to develop apps and to make a new class of devices capable of detecting real emotions.

Gray area; intruding privacy:

Many voices have also been raised on various platforms against the emotions detecting software. Protesters say it is tantamount to intruding into the private life of people and companies should take prior permission from their clients before using the software on them.

Questions have also been raised on the accuracy of results and findings as it could end up making arbitrary and discriminatory decisions.

No comments:

Post a Comment