The technology world is a fast-paced environment, an industry constantly on the move. Over the past few years, wearable gadgets have created a big boom in the market at an astronomical rate and now a smartwatch app has been developed at MIT (Massachusetts Institute of Technology) that detects people’s emotions just by listening to conversations.
Yes, Valentine’s Day is around the corner, so now’s your time to grab a pen and paper and scribble a few notes down. We’ve all been there – gazing across the table into someone’s eyes, either making small talk or engaging in a deep conversation about life and putting the world to rights. And we’ve all worked up a sweat, concerned what we’re saying is not remotely of interest to the person to whom we are speaking.
What Does the Smartwatch App Do?
MIT scientists have developed a prototype app that can independently judge how a conversation is going, and not a moment too soon for anyone who can’t work out why they never seem to get a second date.
(Image source: CNET.com)
You won’t have to practise your one-liners in the mirror prior to setting off any more. When this app is eventually released, you can just go with the flow, be who you are, and this beauty will tell you how you are doing. You can thank Geeknaut later.
How Does it Work?
With artificial intelligence installed, an algorithm within the smartwatch app can sense how you’re feeling in real-time, based on the way you speak. It can tell if you’re happy, sad, boring, awkward or neutral.
The software – downloadable on smartwatches and smartphones – captures data from physiological activity, such as blood flow and pressure, heart rate and skin temperature. It also monitors your tone of voice, including its pitch and energy, and stores text transcripts to assess your use of vocabulary.
In addition, the device takes into consideration the emotions you give off through such ‘tells’ as pausing and fidgeting, and feeds back on how happy or unhappy you seem. This means you effectively have someone else letting you know how you’re coming across. It’s clearly a win-win, no brainer for all parties.
In the below video, this software – which claims 83% accuracy – has been demonstrated on a Samsung Simband:
How Else Could it be Used?
It is hoped that future versions could become a ‘social coach’ for people with anxiety, Asperger’s and other related conditions, as well as detecting excited, tense and boring tones.
The MIT team has also spoken of future improvements to enhance the algorithm’s accuracy. These include being able to monitor whether a conversation is tense and/or exciting, rather than just positive or negative.