Greta Furhat Interface

This page explains

• how to integrate Greta Furhat Interface module to the existing configurations of Greta plateform
• how to run kotlin-coded Furhat Skills from IntelliJ
• how to run the skills on the physical Furhat robot

Requirement

Assuming you can run Greta plateform correctly, make sure the following requirement are satisfied:

• Have the SDK running if you are going to use virtual Furhat
• have Java 8 JDK or OpenJDK 8 installed on your system.
• have a configured microphone.
• have IntelliJ IDEA installed and running. Community edition is enough

Setup: when you launch first time

Note that you can start with Step1 or Step2. But since activemq servers are in Greta, it may be better to start with Step1. If you commence a step, please follow the order

Step 1: Launch Greta Furhat Interface module from modular in Netbeans:

1. Run main project Modular.jar
2. In Greta configurations folder, choose a config that has module "Face KeyFrames Performer", or simply "Face". Ex: Greta-Demo configuration - CereProc.xml
3. In Modular.jar, add Greta Furhat Interface from [Add -> Network Connections -> Greta Furhat Interface -> GretaFurhatAUSenderGui]
4. Add the following Connectors: 4.a. "Face" (or any Face keyframes performers) -> GretaFurhatAUSenderGui 4.b. "Face Viewer Library" -> GretaFurhatAUSenderGui. Note that this connection is necessary only if you are going to use the AUs slider to modify the facial expression of the character
5. Once the module added, click on it to see the User Interface. Then select "Activate" and "Send" to send the AUs on the ActiveMQ topic. Do not change the default values of the server!!! If you do so, make sure to update in Furhat Skills kotlin code

• Save the configuration for future use of the module

• Find below an exemple configuration where "Greta Furhat Interface" is used

• Note: In Greta Furhat Interface, the head movement rotation angles and speech text are directly sent on their respective topic without any need to do something. Their code are integrated in MPEG4Agent.java at auxiliary/OgrePlayer and CereProcTTS.java at auxiliary/TTS/CereProc

Step 2: Launch Furhat Skills in IntelliJ

1. Open up IntelliJ IDEA on the same computer as Greta plateform (necessary for activeMQ servers to work, they are local) and import the skill by going to File->New->Project from Existing Sources -> Import project from external source (also select "Gradle" for project build) -> Create.

   Note: Find the skill GretaMimic at bin/Common/Data/GretaFurhatInterface

2. Create configuration

   Normally, the configuration file should be re-constructed automatically. But if no configuration is found (or simply cannot connect somehow), then create it with the following procedure:

   • click on "current file" next to run -> Edit configuration
   • Add new configuration -> Kotlin
   • refer to image below (in screenshot section) to fill the parts (replace IP address with your robot IP address) And run the Main.kt to run the skill on robot (make sure Step1.5. is OK, all boxes selected)

3. run Main.kt

   3.a. If you are using SDK (virtual Furhat), simply click on run (make sure Step1.5. is OK, all boxes selected)

   3.b. For Furhat robot:

   • Make sure robot is connected on the same network as Greta plateform
   • do "Edit Run Configurations" in IntelliJ and copy the created Run configuration (in order to have two different configurations, one for SDK and the other for robot)
   • Add the JVM argument -Dfurhatos.skills.brokeraddress= with the IP of your robot
   • run the Main.kt

Launch: when you lauch again

1. Launch saved configuration with module GretaFurhatInterfaceGui
2. Select both Activate and Send
3. Run GretaMimic skills in IntelliJ

Note:

For more details, please find the source code of:

• the GretaMimic skills at GretaFurhatInterface/GretaMimic
• Greta Furhat Interface at auxiliary/GretaFurhatInterface

ScreenShot

Configuration with Greta Furhat Interface

Edit configuration in IntelliJ