At NEON, we explored cutting-edge advancements in volumetric capture and 3D reconstruction, pushing the boundaries of realism and interactivity. By leveraging state-of-the-art capture systems, AI-driven animation, and real-time rendering, we create lifelike digital humans for applications ranging from virtual assistants to immersive experiences. Our experiments focus on refining fidelity, natural motion, and scalability, shaping the future of digital interaction.
My Role: Project Project Lead and Management, Technical software and hardware coordination for production, Producer logistic tasks.
Experiment 01: Photogrammetry and 3D Reconstruction.
My Role: Project Project Lead and Management, Technical software and hardware coordination for production, Producer logistic tasks.
Experiment 01: Photogrammetry and 3D Reconstruction.
This is some of our initial experiments to create 3D Avatars through Photogrammetry. The goal of this process was to create a life-like digital human through the process of 3D to allow more control of the digital human. During the capture process, John and I took turns walking around the subjects, taking photos from all angles (above head, side of the face, and under chin) while walking around in a circle.
The photos were then imported into Reality Capture to construct the 3D models.
Experiment 02: Facial Motion Capture
Using motion capture cameras to create precise facial cloud points. These cloud points serve as reference markers for machine learning, allowing for incredibly accurate replication of complex facial expressions in digital human models. By tracking minute details—such as eye movements, subtle muscle contractions, and lip shapes—facial mocap data enables digital humans to display lifelike emotions and reactions.
Experiment 03: 3D Scanning
We took the models we scanned with a handheld scanner and imported them into Meta Human software for modeling and animation.
The results on Meta Human.
