Core ML & Watson Visual Recognition

  1. Introduction
  2. Requirements
  3. Prepare your data
  4. Register your accounts
  5. Create a new project
  6. Understanding your project
  7. Creating a model
  8. Getting the code
  9. Configuring the application
  10. Testing the application
  11. Deploy app to iOS device
  12. Bonus: Improving your classifier
  13. What next?

Bonus: Improving your classifier


  1. If you finish everything and still have time, you can work on improving the classifier that you built. Use the items provided to test your application. To improve your classifier you can take additional images of the items to be used as training data.
  2. Take as many images as you want of each items. Focus on edge cases that are currently not handled well.
  3. Upload the images to your computer.
  4. Back in the Visual Recognition tooling, select your model and click ‘Edit and Retrain.’
  5. Here you can upload your new training images and retrain the model.
  6. You can click the refresh button in the app to pull down the new model after retraining. This update happens on-the-fly without rebuilding the app!

Previous Next