SIMPLE JEV · VISION LAB

Emotion camera

Make a face. Every second or two, a still from your webcam goes to the classifier with four questions: which emotion, how positive, how intense, and whether a face is visible at all.

Vision model

Loading models…

No API key · 2k context · 4 RPS. Frames are sent one at a time, at most one per second, only while the camera is running. Nothing is stored.

Your camera stays off until you press Start camera.

Frame: –
Round trip: –
Input tokens: –

Waiting for a frame

Start the camera to begin.

Emotion probabilities

Valence–

negativepositive

Energy–

calmintense

Valence over the last 60 seconds

Under the hood · latest response

Start the camera to see the API responses.