Project Details
Description
This research project aims to tackle the challenges of generating high-quality 3D digital media content, focusing on achieving realistic and interactive digital human and head avatar models. The primary
research problem is how to efficiently reconstruct highly detailed 3D models that can be seamlessly integrated into journalism and art technology applications. The project addresses key technical issues in
hardware development, algorithmic design, and AI-driven reconstruction to create advanced digital avatars that meet the demands of high-quality media production.
The project is structured around five key tasks. Task 1 aims to develop a light field stage system for capturing multi-angle, multilight source illumination data to reconstruct highly realistic 3D models. Task 2 focuses on reconstructing high-fidelity human head avatars from monocular video, with an emphasis on capturing intricate details, realistic expressions, and ensuring audio-visual consistency. Task 3 seeks to develop advanced digital human avatars by leveraging generative photogrammetry, deep facial rigging, and scalable garment generation to produce realistic animations and dynamic content. Task 4 focuses on addressing intellectual property concerns by developing a robust digital watermarking framework for protecting AI-generated content (AIGC), ensuring secure ownership and authenticity of digital assets. Task 5 explores practical applications of these technologies in journalism and visual arts, aiming to create immersive and interactive experiences that enhance storytelling and creative expression.
The project integrates UIC's expertise in multicamera system hardware development with HKBU's strengths in AI and algorithmic design to build a high-precision 3D reconstruction platform. This collaboration will support diverse 3D content creation projects, connecting transdisciplinary and cross-campus strengths to address applications in journalism, cultural preservation, and art technology.
Additionally, the project will develop secure copyright mechanisms for AIGC, fostering a sustainable digital content ecosystem.
research problem is how to efficiently reconstruct highly detailed 3D models that can be seamlessly integrated into journalism and art technology applications. The project addresses key technical issues in
hardware development, algorithmic design, and AI-driven reconstruction to create advanced digital avatars that meet the demands of high-quality media production.
The project is structured around five key tasks. Task 1 aims to develop a light field stage system for capturing multi-angle, multilight source illumination data to reconstruct highly realistic 3D models. Task 2 focuses on reconstructing high-fidelity human head avatars from monocular video, with an emphasis on capturing intricate details, realistic expressions, and ensuring audio-visual consistency. Task 3 seeks to develop advanced digital human avatars by leveraging generative photogrammetry, deep facial rigging, and scalable garment generation to produce realistic animations and dynamic content. Task 4 focuses on addressing intellectual property concerns by developing a robust digital watermarking framework for protecting AI-generated content (AIGC), ensuring secure ownership and authenticity of digital assets. Task 5 explores practical applications of these technologies in journalism and visual arts, aiming to create immersive and interactive experiences that enhance storytelling and creative expression.
The project integrates UIC's expertise in multicamera system hardware development with HKBU's strengths in AI and algorithmic design to build a high-precision 3D reconstruction platform. This collaboration will support diverse 3D content creation projects, connecting transdisciplinary and cross-campus strengths to address applications in journalism, cultural preservation, and art technology.
Additionally, the project will develop secure copyright mechanisms for AIGC, fostering a sustainable digital content ecosystem.
| Status | Active |
|---|---|
| Effective start/end date | 1/05/25 → 30/04/28 |
Fingerprint
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.