UnrealGPT was part of my undergraduate thesis project aimed at combining Unreal Engine 5, Virtual Reality (VR), MetaHuman, OpenAI, and Azure TTS to create realistically interactable non-player characters (NPCs) through large language models (LLMs) and speech synthesis. Please note that a portion of the code is currently under academic proprietorship until October 2024.

EDIT: A fully-developed demo of the application is now available here: https://github.com/oscaem/ARIS-Unreal-GPT

The goal of UnrealGPT is to leverage the power of cutting-edge technologies to enhance the immersion and interactivity of NPCs in virtual environments. By integrating Unreal Engine 5, which provides stunning visuals and realistic physics, with VR capabilities, MetaHuman for realistic character models, OpenAI's language models for natural language understanding and generation, and Azure TTS for high-quality speech synthesis, we aim to create NPCs that can understand and respond contextually to their surroundings.

• Unreal Engine 5: We utilize Unreal Engine 5 as the core framework for building immersive virtual environments. It offers state-of-the-art graphics, physics, and rendering capabilities that enhance the realism of the NPCs.

• Virtual Reality (VR): UnrealGPT supports VR integration, allowing users to interact with the NPCs in a more intuitive and immersive manner. VR headsets provide a first-person perspective and enable natural hand and body movements within the virtual environment.

• MetaHuman: By leveraging the power of MetaHuman, we create highly realistic character models for our NPCs. These models possess lifelike features and animations, enhancing the overall visual fidelity and believability of the virtual characters.

• OpenAI Language Models: We integrate OpenAI's language models to enable natural language understanding and generation for the NPCs. This allows them to comprehend and respond contextually to user inputs and dialogue, creating more engaging and interactive conversations.

• Azure TTS: Azure TTS is used for speech synthesis, enabling the NPCs to speak with high-quality and human-like voices. This enhances the realism of the interactions and provides a more immersive experience for the users.

Please note that a portion of the code for UnrealGPT is currently under academic proprietorship through October 2024. During this period, the code is not available for public distribution or modification without explicit permission.

As the code is not publicly available at the moment, the installation and usage instructions cannot be provided in this README. However, if you have any questions or inquiries regarding the project or its implementation, please feel free to contact me directly.

As UnrealGPT is a part of my undergraduate thesis project, external contributions are currently not accepted. However, your feedback and suggestions are highly appreciated and can help shape the future development of the project.

If you have any questions or require support related to UnrealGPT, please don't hesitate to reach out to me via the provided contact information.

For any inquiries or further information about UnrealGPT, please contact:

• Oscar
• [email protected]

Thank you for your interest in UnrealGPT!