Physics-Based Animation

Bazin, Jean-Charles; Plüss (Kuster), Claudia; Yu, Guo; Martin, Tobias; Jacobson, Alec; Gross, Markus

Convincing manipulation of objects in live action videos is a difficult and often tedious task. Skilled video editors achieve this with the help of modern professional tools, but complex motions might still lack physical realism since existing tools do not consider the laws of physics. On the other hand, physically based simulation promises a high degree of realism, but typically creates a virtual 3D scene animation rather than returning an edited version of an input live action video. We propose a framework that combines video editing and physics-based simulation. Our tool assists unskilled users in editing an input image or video while respecting the laws of physics and also leveraging the image content. We first fit a physically based simulation that approximates the object’s motion in the input video. We then allow the user to edit the physical parameters of the object, generating a new physical behavior for it. The core of our work is the formulation of an image-aware constraint within physics simulations. This constraint manifests as external control forces to guide the object in a way that encourages proper texturing at every frame, yet producing physically plausible motions. We demonstrate the generality of our method on a variety of physical interactions: rigid motion, multi-body collisions, clothes and elastic bodies.

Physically Based Video Editing

Matthew Cong, Kiran S. Bhat, Ronald Fedkiw

We propose a new framework for the simulation of facial muscle and flesh that so significantly improves the technique that it allows for immediate mainstream use of anatomically and biomechanically accurate muscle models as a bread and butter technique in a high-end production quality pipeline. The key idea is to create a blendshape system for the muscles that gives the precise directability and controllability required in a high-end production environment. The blendshape muscles are used to drive the underlying anatomically and biomechanically motivated simulation in a way that is unbound by the typical restrictions of a simulation system while still retaining the desirable degree of freedom richness that leads to high quality results. We show that we are able to target production quality facial shapes, whether from scans or an animation system, and illustrate that the resulting nonlinear simulation in-betweens are of higher quality than those obtained from traditional linear blendshapes. We also demonstrate the ability to selectively improve areas on a given blendshape using the results of a simulation, as well as the ability to edit muscle shapes and paths in order to produce directability for animator control. Then, we show how these techniques can be used to transition from one blendshape to another or even track and selectively modify an entire performance. The efficacy of our system is further demonstrated by using it to retarget animation onto new creature models given only a single static rest pose as input.

Art-Directed Muscle Simulation for High-End Facial Animation