Helper scripts to quickly install YOLO darknet image recognition neural network into a linux shell.

It can take snapshots of the desktop screen, the system webcam, or from an Ambarella wifi streaming live camera.

This will detect objects into snapshots and speak them with espeak.

As it use NO GPU, it can run pretty anywhere.

The snapshots and recognition process takes around 1-2 seconds, depending the running machine.

Dev goal: Automotive usage.

Usage: ./scan [[url|screen|webcam|cam] [en|fr]

Default: webcam

./install

Or directly:

php <(curl https://webdev23.github.io/aten/install)