This class will focus on developing computational models of human movement. As part of this course, we will develop algorithms and techniques for animating and simulating embodied autonomous virtual humans, spanning approaches for humanoid character animation, algorithms for navigation and way finding, goal-directed collision avoidance, and multi-actor behavior authoring. We will discuss how these approaches can scale to create simulate human crowds (with hundreds and thousands of virtual humans). A special emphasis will be placed on autonomy (artificial intelligence) and interactivity.
Course Topics (Tentative)
- Curves
- Collision Avoidance
- Particle Dynamics
- Crowd Simulation
- Social Forces for Goal-Directed Collision Avoidance
- Reciprocal Collision Avoidance
- Keyframe Animation
- Motion Capture Techniques for Humanoid Character Animation
- Motion Graphs
- Procedural Character Animation
- Rigid Body Dynamics
- Navigation and Pathfinding
- Behavior Authoring
- Parameterized Behavior Trees
- Computer-Assisted Narrative Animation Synthesis
- Computer-Assisted Authoring of Interactive Narratives
Class Logistics
- Heavily programming oriented course
- Programming in C++ and C#
- > 10 assignments
- No written exams.
- Will work on large "industry standard" code bases
- Will work in groups
Special Permission Numbers
SPN’s will be handed out closer to the beginning of class. Please use the online portal to submit an SPN request