Lecture 11 · Parametric Modeling Basics
Algorithmic Design: Physically-based Modeling; Generative AI
Lecture 11 Physically-based Modeling and Generative AI
-
Physically-based Modeling
Examples of Kangaroo Physics -
Homework 6 - Generative AI to 3D Modeling
Review the OpenAI Sora text-to-video example and the video-to-3D model example below:- OpenAI Sora example (Amalfi Coast)
- OpenAI Sora to 3D Gaussian Splatting example (Amalfi Coast)
Watch more AI-generated videos and read their prompts from https://openai.com/sora
Use any 2 videos of your choice downloaded from: https://openai.com/sora
and the 3D reconstruction tool on https://poly.cam/
(free trial version) -> Create New Capture -> Create a Gaussian Splat,
to generate 3D models for the videos.
(Remember to cancel your trial version within 6 days to avoid paying if you don't want to keep it.)
For better 3D modeling results, try to use videos with clear 3D objects and with multiple views of the 3D objects.
Post your models' URLs in Homework 6 on Canvas, so that the instructor can check the results.
Think of the differences (e.g., about how much they support editing, parametric changes, and their applications) between a 3D model constructed this way and a parametric 3D model that we can create using Rhino/Grasshopper.
Optional: If you have access to a VR headset, try to experience the AI-generated 3D scenes in VR!
Submission
| Item | Details |
|---|---|
| Information to submit | Post URLs of 3D models to Homework 6 on Canvas |
| Submit to | https://canvas.tamu.edu/ |
| Due | 2/28, Wednesday, 10:00 PM |