Goals
This VIP project aims to create physically-based neural graphics algorithms that can generate visually appealing computer animations depicting various natural phenomena and physical processes. Our research is at the intersection of computer graphics, computational physics, neural representation, and machine learning.
Issues Involved or Addressed
Compared to the extensive time and computational resources once required for intricate details like Merida's hairstyle or the ocean animations in Pirates of the Caribbean, emerging AI-powered graphics algorithms are revolutionizing the field, making advanced visual effects more accessible to everyday content creators than ever before. This VIP project aims to invent novel algorithms by bringing together the best of two worlds: traditional, physics-based simulation methods and the fast-evolving field of AI-driven dynamic graphics. The mission of this project stands at a unique intersection of multiple disciplines, from computer graphics and computational physics to machine learning and data science, presenting a unique opportunity to explore the best practice of integrating physical authenticity into AI-generated graphics. By developing novel data structures, PDE solvers, neural representations, and machine learning paradigms, we aim to significantly elevate the state of the art in physically-based animation and pave the way for innovations that could revolutionize how we create and interact with the physical world.
Methods and Technologies
Academic Majors of Interest
- Computing›Algorithms, Combinatorics and Optimization
- Computing›Analytics
- Computing›Computational Science and Engineering
- Computing›Computer Science
- Computing›Human-Centered Computing
- Computing›Human-Computer Interaction
- Sciences›Applied Physics
- Sciences›Biology
- Sciences›Mathematics
- Sciences›Physics
Preferred Interests and Preparation
Familiarity with Python and C++ is preferred, along with a basic understanding of graphics, physics, and machine learning.
Meeting Schedule & Location
Team Advisors
- College of Computing