This project provides a Live2D virtual human for chatting, built with Unity. It targets users who want to create interactive AI-powered virtual companions, offering a visually engaging and conversational experience.

How It Works

The system integrates Unity for rendering Live2D models and handling user interaction. It leverages external APIs, specifically Azure and OpenAI, for natural language processing and conversational AI. Image processing for face detection is handled by OpenCVPlusUnity, and JSON operations use Newtonsoft.Json.

Quick Start & Requirements

• Environment: Unity 2021.3.0.
• Prerequisites: API keys for Azure, OpenAI, or APISpace are required.
• Setup: Users need to import specific Unity packages (OpenCVPlusUnity, Newtonsoft.Json) and configure API keys within the application's UI.
• Documentation: Technical documentation is available for development insights.

Highlighted Details

• Live2D model rendering within the Unity engine.
• Integration with external AI services (Azure, OpenAI) for conversational capabilities.
• Utilizes OpenCV for image processing tasks like face detection.
• JSON serialization/deserialization via Newtonsoft.Json.

Maintenance & Community

The project is open-source under the MIT license. Acknowledgements are made to specific artists for model and background assets, and to an individual for providing an OpenAI API proxy for domestic network environments.

Licensing & Compatibility

• License: Code is licensed under MIT.
• Assets: Live2D models and art resources are owned by their respective authors and must be used under their individual licenses.
• Commercial Use: Compatibility for commercial use depends on the licenses of the included art assets and the terms of service for the AI APIs used.

Limitations & Caveats

The project does not provide public API keys and requires users to obtain their own. The licensing of art assets may impose restrictions on commercial use or redistribution.