This project is for UI task automation using multiple "windows". The windows are a transparent overlay that can capture the pixels on the screen. For example, if one wanted to press a button when the color of the button was green place a window over the button and use some computer vision methods to determine what the window is seeing.
The whole premise is to make the viewports as small as possible on screen because the larger the resolution supplied to a Convolutional Neural Network the more complex the model has to be.
TBF
This project uses the Tensorflow JNI. The libtensorflow.jar and JNI files can be downloaded from the link below. https://www.tensorflow.org/install/lang_java