Track Emotions For A Deeper Understanding

Transforming conversations using neuroscience, artificial intelligence, computer vision, and mathematics.

30-day trial. No credit card required. Access all features.

Using only the end-users camera. All data remains in the browser. 

How we detect emotions using camera-only

Immersively.care integrates world-leading emotion detection engines directly in the browser with no server-side analysis.

1. Face detection

Each frame of the video stream is processed inside the user's browser. The largest face in front of the camera is detected using machine learning. The photos are never transmitted, stored, or shared.

2. Facial features extraction

Immersively.care extracts facial landmark positions from each facial image, such as the end of the nose or eyebrows, the eyes, and mouth, as well as groups of facial muscles and changes in those positions.

3. Classification of facial features as emotion

  • We employ deep learning that teaches our algorithms the likelihood of facial features being related to seven core emotions – anger, disgust, fear, happiness, sadness, and surprise, plus the baseline neutral expression.
  • The algorithms have been pre-trained using large custom datasets with huge amounts of faces, ranging from various geographies, ethnicities, ages, and genders, labeled with the emotions that individuals have experienced.
  • During a video call, Immersively.care continuously executes the algorithms on each facial image or succession of images in real-time and displays the three emotions which have the highest probability of prediction. 

 

Happiness

satisfaction, contentment,

joy

Sadness

disappointment, grief,

hopelessness

Fear

anxiety, apprehension,

nervousness

Anger

hostility, agitation,

frustration

Surprise

unexpected, shock,

astonishment

Disgust

disapproval, unpleasentness,

contempt

Immersively.care loads emotions recognition engines inside the user's browser only upon user consent.

Emotion detection requires facial landmark positions only, which are just a bunch of numbers.

Therefore the original photo of the face is not stored, and never leaves the user's computer. All facial data is processed anonymously on the user side in the browser and immediately destroyed within a few milliseconds.

What Researchers Are Saying

"Immersively.care's application of facial video-based frameworks to estimate physiological data using Remote Photoplethysmography is a revolution in the field of virtual meetings. I tested it with my Smart Watch on my wrist. Give it a try!"

Mark S.Mark S.Assistant Professor

"I was quite skeptical at the beginning, because of the lack of scientific consensus on the definition of emotions. Immersively.care found a way to ethically offer emotions as an additional language of communication during a video call. Great job!"

Sarah B.Sarah B.Chief Psychologist

"I like how the technology doesn't try to read people's minds but instead seeks to copy the human action of looking at someone's face to guess how they're feeling. This extra context can be helpful during a remote call"

Roseta S.Roseta S.Accessibility Researcher

Ready to give it a try?

Onboarding our technology is extremely fast and easy.

Step 1 - sign up and log in
Step 2 - click on "Meet Now"
Step 3 - invite your guest

DigIntu Tech Ltd

Hermanova 61

17000 Prague

Czech Republic

EU VAT-ID CZ05233518

Registered in Commercial Register with the Municipal Court in Prague, section C, folder 260406.

Sign up for newsletter

Thank You, we'll be in touch soon.

FEATURES

Accessibility support

Instant messaging

Live transcription

Live translation

Scheduling

Journaling

Telehealth

 

COMING

Virtual backgrounds

Stress level tracking

Blood pressure tracking

Oxygen level tracking

Breathing rate tracking

Copyright ©Immersively.care 2023