Webcam Style Transfer using NFST

Models from (https://github.com/microsoft/Windows-Machine-Learning/tree/master/Samples/FNSCandyStyleTransfer/UWP/cs/Assets) , webcam using onnx, cinder framework and opencv, inference only no training.

Before building:

• Reinstall Microsoft.ML.OnnxRuntime.DirectML from Nuget package manager
• Update the 'include' and 'library' paths for OpenCV (from project settings).

If different modes are used, input and output need to be changed accordingly

A different example using face emotion detector model is in the other branch of the this same repo

nfst_demo.mp4