Start-up's AI can read human emotions, even in large crowds

The Manila Times

Start-up's AI can read human emotions, even in large crowds  Spanish technology startup, Neurologyca, says its Kopernica artificial intelligence (AI) system can discern how you feel or detect a range of neurological conditions, and is even capable of doing this simultaneously for a crowd of people.  Neurologyca hopes to advance emotional intelligence in AI, enhancing the way machines understand and interact with humans.  Kopernica provides real-time emotional intelligence assessments of anyone on its video feed, aiming to enhance human-machine interactions.  Kopernica uses multi-modal intelligence, including Natural Language Processing (NLP), voice recognition, body language interpretation, and personality data analysis, to enhance its computer vision and provide a holistic understanding of an individual's emotional state; and even vast crowds of people.  REUTERS VIDEO  Subscribe to The Manila Times Channel - https://tmt.ph/YTSubscribe  Visit our website at https://www.manilatimes.net  Follow us:  Facebook - https://tmt.ph/facebook  Instagram - https://tmt.ph/instagram  Twitter - https://tmt.ph/twitter  DailyMotion - https://tmt.ph/dailymotion   Subscribe to our Digital Edition - https://tmt.ph/digital   Check out our Podcasts:  Spotify - https://tmt.ph/spotify  Apple Podcasts - https://tmt.ph/applepodcasts  Amazon Music - https://tmt.ph/amazonmusic  Deezer: https://tmt.ph/deezer Stitcher: https://tmt.ph/stitcher  Tune In: https://tmt.ph/tunein  Soundcloud: https://tmt.ph/soundcloud   #TheManilaTimes #tmtnews #artificialintelligence

Transcript

00:00Well, we are we are a company where we spend many, many years

00:29training our models based on emotions and cognitive states

00:33for measuring the emotions and cognitive state of people, and

00:36also the mood of people.

00:51So there's your face. And we're looking at we're looking at

00:54levels of joy, trust, anticipation, we can cycle

00:58through up to 90 different types of emotions and get a real time

01:02input on what you're feeling right now is we're having this

01:05conversation.

01:20Our technology uses a library of over 13 billion data points to

01:26map human emotion to video signals. And so what you see

01:30here are three different cameras measuring three different

01:34things. This first camera is what we call our massive

01:37measurement. So all of these little dots you see here are all

01:41the people sitting over there. And so we can actually measure

01:45hundreds and even 1000s of people at a time and get the

01:48general sentiment of a crowd. Our technology is installed in

01:52soccer stadiums and airports. And we can determine whether a

01:56crowd is interested, enthusiastic, bored, excited,

02:02and either dynamically change advertising, or in the case of a

02:06soccer stadium, determine whether the crowd is getting so

02:09rowdy that somebody needs to intervene.

02:11Our mission here is just to try to humanize AI to make AI more

02:30human out, make the AI understand much better the

02:34communication with a human being from the emphatic point of

02:37view, from the emotional point of view from the subjective

02:41point of view.

03:07So this is a heat map of all of the different emotions that

03:10this person's experiencing.

03:20This is a single face measurement in pretty high

03:22fidelity. You can see here there's this one's measuring

03:25stress, intensity and positivity, we can measure up to

03:2992 different emotions, this is just showing three of them. And

03:33those are used to determine everything from health and

03:38wellness applications, to public safety, to gaming.

03:48It turns out that when somebody is more likely to have a stroke

03:54or brain hemorrhage, they start exhibiting unique emotional

03:59patterns that are imperceptible to the human eye, but that we

04:02can capture with our technology. And there are micro movements of

04:05your face, there are dilations of the pupil. And so we can

04:09predict weeks before it happens, an event like that.

04:21Yeah, so it's these dots here that show the faceless

04:24capability of the platform. So if you see here, there's little

04:28purple markers, these little purple markers move around. So

04:33our system is completely faceless. So we never capture

04:36any personally identifying data. And all of this data is mapped

04:41to an arbitrary number that is constantly changing. So it was

04:45built in Spain, complying to all GDPR and privacy standards. You

04:52can measure someone's face for things like education and

04:54clinical applications. But you can also turn that off and have

04:58it be completely faceless.

Category

Transcript

Recommended