Installation

The InteractML Plugin is available for free on the 🔗Unreal Marketplace.

1. Open the Epic Games Launcher
2. On the Marketplace tab search for InteractML
3. Go through the 'add to cart' and 'purchase' procedure (it's free)
4. In the Library tab
5. Locate the InteractML plugin in the Vault section
6. Click the "Install To Engine" button
7. Select the Unreal Engine version you want to use it with
8. Click Install
9. Once installation is complete you are ready to use it in your projects

Project

Enabling the plugin for use in a new or existing project is straightforward:

1. Create new project / open existing project
2. Open the Plugins window (Edit menu --> Plugins)
3. Select the Machine Learning sub-section of the Installed plugins
4. Click the Enabled checkbox on the InteractML plugin entry
5. Accept the "Beta Version" warning popup (Yes)
6. Click the Restart Now button at the bottom of the Plugin window (adding plugins requires a restart)
7. Close the Plugins window
8. You are now ready to use InteractML!

Minimal Viable Example

It's fairly easy to set up a minimal viable test case as follows:

1. Create a new InteractML Training Set asset
   1. Right click in Content Browser
   2. Select Training Set from the Interact ML submenu.
   3. Rename it (e.g. MyTrainingSet)
2. Create a new InteractML Classification Model asset
   1. Right click in Content Browser
   2. Select Classification Model from the Interact ML submenu.
   3. Rename it (e.g. MyModel)
3. Locate the Character Blueprint you are using
   1. E.g. in the First Person Game Template it's in FirstPersonBP/Blueprints/FirstPersonCharacter
   2. Open the blueprint (by double clicking on it)
4. Add some extra variables to the Blueprint
   1. A float variable called "ExpectedOutput"
   2. A Boolean variable called "Recording"
   3. A Boolean variable called "Training"
5. Ensure you are on the Event Graph tab
6. Add a node graphs like the ones shown below (type "InteractML" in the node menu to find the custom nodes)
7. Set the Training Set input of the Example Recorder node to the Training Set asset you created above
8. Set the Model and Training Set inputs of the Teach The Machine node to the assets you created above
9. Set the Model input of the Machine Learning Robot to the Model asset you created above
10. Click Save and Compile
11. Press Play to test it

Instructions

1. Press 'R' to record a player position
2. Press '4' and '5' to change the current Expected Output value
3. Record some more locations with different Outputs (training and running is automatic)
4. Revisit the recorded locations and the output displayed at the top-left should change

Graphs

The main Recording, Training, and Running functionality.

images/ui/Example_MinimalMachineLearning.png

The user interaction functionality.

images/ui/Example_MinimalInteraction1.png

images/ui/Example_MinimalInteraction2.png

Source

If you require access to the full source for the plugin and example projects have a look at 🔗the InteractMLUE repository.

---

👈 [Team and Terms]] ](/Interactml/iml-unreal/wiki/🏠-[[Home) | Machine Learning 👉