Vision telematics

.
Extract live data and insights from your camera streams.

Your workspace

Add pictures or a short video with the plus button.

0 of 0

Advanced

These are my picks for this job. Change anything and it stays yours.

Model size

Resolution

Passes over your pictures

Pictures at a time

Stop early after

Held back for marking

Confidence to call it

Overlap before merging

Lighting copies per picture

Light and dark jitter

Rotation

Zoom jitter

Mirror left to right

Collage four together

I marked this one. Does it look right?

Drag a box around it, or let me mark them all and just check my work.

Getting your pictures ready

0% working out how long this will take

Sending your pictures
Getting the server ready
Learning what your thing looks like
Packing the model up
Putting it to work

You can close this page. Training carries on and the model will be waiting when you come back.

Scan this with the phone you want to use as the camera.

Waiting for the phone…

Confidence 30% Overlap 45% changes apply to the very next frame

Solid boxes clear your confidence bar. Faint dashed ones are what the model ALMOST called, so a model that is close but shy is visibly different from one that sees nothing.

Confidence 30% Overlap 45% this view now; a linked phone follows within ten seconds

This is what the camera sees right now.

Your prompt

Lightapp is building your application

Understanding your prompt
Connecting your camera
Choosing what to look for
Getting this device ready
Your application is ready

Everything runs on this device. The first start takes a few seconds.

.
Extract live data and insights from your camera streams.

Your conversations

Free plan

Sign in to build

Hi there.Extract live data and insights from your camera streams.

Your conversations

.
Extract live data and insights from your camera streams.