MakeML is a Developer Tool for Creating Object Detection and Segmentation Neural Networks without a Line of Code. It's built to make the training process easy to setup. It is designed to handle data sets, training configurations, markup and training processes — all in one place.

MakeML Nails

Is an iOS app example that shows how users can try on different nail polish colors with semantic segmentation. The machine learning model was trained for 3 hours with 50 photos. Have used tensorflow for training model and their SDK for iOS.

Train Objects Segmentation .tflite model

See the Tutorial for the training object segmentation model without a line of code with macOS desktop application.

Using another .tflite model in iOS application

For using MakeML Nails project with another .tflite file, add it to the project and change this line with your name of the model.

NSString *modelPath = FilePathForResourceName(@"result_model", @"tflite");

Change the number of classes and add colors to the array:

int class_count = 2;
unsigned int colors[class_count] = {  0000000000, additionalColor };

Links

More Tutorials | MakeML in App Store | Full Documentation | MakeML Chat | Support page