🚀 Omniverse & Melax Setup Guide

🎮 Omniverse Installation & Configuration

1. Download Omniverse Launcher

   • From: wingetgui.com

2. Create & Log in to NVIDIA Account

   • Sign up or log in as required.

3. Set Path Configuration

   Library Path: C:\Users\YourName\AppData\Local\ov\pkg
   DATA Path: C:\Users\YourName\AppData\Local\ov\data
   Content Path: C:\Users\YourName\Downloads
   CACHE Path: C:\Users\YourName\AppData\Local\ov\cache

4. Launch Omniverse

5. Install Audio2Face

   • Navigate to Exchange → Search for Audio2Face
   • Download version 2023.1.1

6. Run Audio2Face in Headless Mode

   • Edit start_headless.bat with the correct path:

     set TARGET_BATCH="C:\Users\YourName\AppData\Local\ov\pkg\audio2face-2023.1.1\audio2face_headless.bat"

---

🧠 Melax Setup

1. clone from https://github.com/hcmlab/melax.git

2. Run Installer Download Cora

   • Download the character developed in Unreal Engine 5.3: https://uniaugsburg-my.sharepoint.com/:f:/g/personal/daksitha_withanage_don_uni-a_de/EnzWciX1_OJHkfvTXJhz4Z0BQl1aWLAdLkRg-tJZsKZ04w?e=PE9Ao0 The GUI and Audio2Face server must be running.
   • Execute MeLaX_ACE_A2F_Only.exe

3. Network Permissions

   • On first run, allow access on Public, Private, and Domain networks.

4. Download Ollama

   • Ensure it is installed on your system. https://ollama.com/download/windows

---

🐍 Running mainwindow.py

1. Install Python Packages

   • From requirements_python3_11.txt (for Python 3.11)

2. Install ffmpeg if not already installed

3. Run the GUI

   python gui.py

---

Once everything is running: In the GUI load the a2f model: select 'female.usd' and press load button. Select your microphone and choices of AST, LLM, TTS. The press play button on the GUI.

You should be able to speak to the character in an interactive manner.