This project focuses on building an AI-powered digital human chatbot designed to help patients and visitors navigate a hospital platform. The goal was to create a realistic, responsive virtual assistant capable of answering user queries, providing guidance, and enhancing the overall digital healthcare experience.

iClone

I discovered iClone as an emerging 3D animation software and leveraged its:
1. Facial mocap capabilities
2. Lip-sync improvement tools
3. Character motion editing features

Unreal Engine (Metahumans)

This became the core tool due to:
1. High-fidelity human models
2. Real-time animation system
3. Easy character customization

Adobe Mixamo

Used for initial rigging and testing character animations.

Blender

Used for refining models, adjusting textures, and custom body movement adjustments.

Traditional hospital web and kiosk systems often feel static and impersonal. Users—especially first-time visitors or elderly patients—require a more intuitive and human-like guide.

A lifelike virtual assistant with natural gestures.

Real-time speech interactions.

Accurate lip-sync and facial expressions.

Easy-to-edit character models for future hospital use cases.

We designed a 3D digital human assistant using a combination of advanced animation and AI technologies.

Human-like body movements, eye blinking, and gesture animations.

Real-time speech output with prompt-driven responses.

Customizable avatars based on Unreal Engine Metahumans.

Integration with AI chatbot logic for conversational flow.

After nearly 50 days of development, we achieved a fully functional digital human assistant and The final output meets hospital requirements and serves as a scalable foundation for future AI-driven virtual assistants in healthcare.

Natural facial expressions.

Smooth gestures and blinking.

Accurate lip-sync.

Real-time conversational capabilities.

Clean, customizable character design.