No face yet
Press Start and look at the camera.
Face tracking runs in your browser. Photos and data are not uploaded. They only leave this page if you download the PDF or CSV.
Not a medical diagnosis. Always consult a doctor. This tool reads facial expressions, not feelings, and a camera can get it wrong. Use the results only as a conversation starter with a qualified doctor or licensed mental health professional.
2-minute emotion test
Eight prompts, 15 seconds each. Read each one on the camera view and react naturally. The first is a neutral baseline. A few prompts are mildly sad or irritating, and the test ends on a positive note. You can stop at any time, and it should not be used by anyone who is feeling distressed.
| # | Prompt | Target | Peak | Reaction | Result |
|---|
Observed affect (last 30 seconds)
Collecting data…
Press Start and keep your face in view for about 5 seconds.
Indicators
Whole session
- Active time
- 0:00
- Face in view
- –
- Readings
- 0
Report
How the numbers are worked out
- About 5 readings per second, each with four values from 0 to 1: angry, sad, surprised, happy.
- Each reading is placed on a map of mood (unpleasant to pleasant) and energy (low to high). Happy is pleasant with medium energy, sad is unpleasant with low energy, angry is unpleasant with high energy, surprised is neutral with high energy.
- The label uses the average position over the last 30 seconds. The clinical terms come from the mental status examination vocabulary (affect, range, intensity, stability) and describe what the face looks like, not a diagnosis.
- Positivity is the average happy value. Tension is angry plus half of surprised. Low mood is the average sad value. Steadiness is how little the mood position moved.
- An emotion counts as shown when it passes 0.4. Otherwise the reading is counted as neutral.
- In the emotion test, a prompt counts as Responded when its target emotion passes 0.4, and Mild response when it passes 0.25 or rises by 0.15 over the baseline. Reaction time is how long it took to pass 0.4.
- The thresholds are reasonable starting values and have not been clinically validated.