Using a built in model with CoreML

Question

Is it possible to use an already built in model (e.g. Food) from the Watson Studio in your CoreML app? I can't figure out how to do so if it is. All the documentation there is tells me to train my own custom model, but I want to use the already existing food model.

This is what I have done so far:

let apiKey = "xyz"
    let classifierId = "food"
    let version = "2017-12-07"
    var visualRecognition: VisualRecognition!

    override func viewDidLoad() {
        super.viewDidLoad()
        self.visualRecognition = VisualRecognition(apiKey: apiKey, version: version)
    }

    override func viewDidAppear(_ animated: Bool) {
        // Pull down model if none on device
        let localModels = try? visualRecognition.listLocalModels()
      print(localModels)
        if let models = localModels, models.contains(self.classifierId)  {
            self.currentModelLabel.text = "Current Model: \(self.classifierId)"
        } else {
            self.invokeModelUpdate()
        }
    }

But running this gives me a 404: Status code was not acceptable: 404

Chris Chris · Accepted Answer · 2018-04-12T17:29:52

This is not possible. I got a response back from the IBM team. Hope this helps someone else.

Using a built in model with CoreML

1 Answers