Large Set of Pre-Trained Neural Nets for Images
The Wolfram Neural Net Repository hosts an expanding collection of neural net models that work on images for immediate evaluation and includes a large number of networks that work.
Here is an example that segments an image into semantic concepts.
Write an evaluation function to handle net reshaping and resampling of input and output.
Define the labels associated to each class number.
Run the evaluation function and detect the available classes.
Colorize the resulting segmentation.