Goto

Collaborating Authors

 mlfeatureprovider


How To Train a Core ML Model on Your Device

#artificialintelligence

MLModel is the class that encapsulates the model. The Core ML model accepts the inputs and outputs in the form of MLFeatureValue. MLFeatureValue lets us directly use a CGImage. Along with that, we can pass the image constraints for the model. It creates the CVPixelBuffer from the CGImage for you, thereby avoiding the need to write helper methods.