Computer Graphics
| Disney Research has a strong competency in computer graphics. Our entertainment businesses provide diverse target applications for our pioneering work. Because of this we achieve a rare level of cross-fertilization by juxtaposing real-time algorithms for the game studios with high-end techniques for the movie studios, achieving speed and directability in physical simulation, spanning visual styles from photorealistic to artistic, and blurring the boundaries between computer graphics and materials science. |
In this work, we present a method for generating animations of non-humanoid characters from human motion capture data. Characters considered in this work have proportion and/or topology significantly different from humans, but are expected to convey expressions and emotions through body language that are understandable to human viewers. Keyframing is most commonly used to animate such characters. Our method provides an alternative for animating non-humanoid characters that leverages motion data from a human subject performing in the style of the target character. The method consists of a statistical mapping function learned from a small set of corresponding key poses, and a physics-based optimization process to improve the physical realism. We demonstrate our approach on three characters and a variety of motions with emotional expressions. [
More...]
Linear models, particularly those based on principal component analysis (PCA), have been used successfully on a broad range of human face-related applications. Although PCA models achieve high compression, they have not been widely used for animation in a production environment because their bases lack a semantic interpretation. Their parameters are not a natural set for animators to work with. In this paper we present a linear face modelling approach that allows intuitive click-and-drag interaction for animation. [
More...]
Modular Radiance Transfer
Modular Radiance Transfer is an approach for interactively computing approximate direct-to-indirect transfer by warping and combining transport from a library of simple shapes. Incorporating precomputed light transport into authoring pipelines for large scenes incurs long preprocessing times, generates large datasets, hinders artistic iteration workflows, and often results in only modest run-time performance. We observe that using a prior on the distribution of incident lighting enables accurate low-rank approximations to the light transport operator for simple canonical shapes, which can be precomputed off-line. An implicit lighting environment induced from the low-rank approximation is then used to model the flow of light volumetrically in the scene and through interface lightfields between shapes. These interfaces enable coupling between shapes and act as aggregation points for distant propagation, increasing the runtime performance and minimizing the required memory. We replace the scene dependent precomputation with a light-weight, artist driven mapping between the complex scene and the dictionary of shapes. High frame rates are produced on target platforms ranging from cell-phones to high end GPUs [More...]
Although animation is one of the most compelling aspects of computer graphics, the possibilities for depicting the movement that make dynamic scenes so exciting remain limited. In our work, we experiment with motion depiction as a first-class entity within the rendering process. We extend the concept of a surface shader, which is evaluated on an infinitesimal portion of an object's surface at one instant in time, to that of a programmable motion effect, which is evaluated with global knowledge about all portions of an object's surface that pass in front of a pixel during an arbitrary long sequence of time. [
More]
We present a method for generating art-directable volumetric effects, ranging from physically-accurate to non-physical results. Our system mimics the way experienced artists think about volumetric effects by using an intuitive lighting primitive, and decoupling the modeling and shading of this primitive. To accomplish this, we generalize the physically-based photon beams method to allow arbitrarily programmable simulation and shading phases. This provides an intuitive design space for artists to rapidly explore a wide range of physically-based as well as plausible, but exaggerated, volumetric effects. We integrate our approach into a real-world production pipeline and couple our volumetric effects to surface shading. [
More...]
We present two contributions to the area of volumetric rendering. We develop a novel, comprehensive theory of volumetric radiance estimation that leads to several new insights and includes all previously published estimates as special cases. This theory allows for estimating in-scattered radiance at a point, or accumulated radiance along a camera ray, with the standard photon particle representation used in previous work. Furthermore, we generalize these operations to include a more compact, and more expressive intermediate representation of lighting in participating media, which we call ``photon beams.'' The combination of these representations and their respective query operations results in a collection of nine distinct volumetric radiance estimates.
Our second contribution is a more efficient rendering method for participating media based on photon beams. Even when shooting and storing less photons and using less computation time, our method significantly reduces both bias (blur) and variance in volumetric radiance estimation. This enables us to render sharp lighting details (e.g. volume caustics) using just tens of thousands of photon beams, instead of the millions to billions of photon points required with previous methods. [More...]
Stable Spaces for Real-time Clothing
We present a technique for learning clothing models that enables the simultaneous animation of thousands of detailed garments in real-time. This surprisingly simple conditional model learns and preserves the key dynamic properties of a cloth motion along with folding details. Our approach requires no a priori physical model, but rather treats training data as a 'black box'. We show that the models learned with our method are stable over large time- steps and can approximately resolve cloth-body collisions. We also show that within a class of methods, no simpler model covers the full range of cloth dynamics captured by ours. Our method bridges the current gap between skinning and physical simulation, combining benefits of speed from the former with dynamic effects from the latter. We demonstrate our approach on a variety of apparel worn by male and female human characters performing a varied set of motions typically used in video games (e.g., walking, running, jumping, etc.).
Upsampling for Cloth
We propose a method for learning linear upsampling operators for physically-based cloth simulation, allowing us to enrich coarse meshes with mid-scale details in minimal time and memory budgets, as required in computer games. In contrast to classical subdivision schemes, our operators adapt to a specific context (e.g. a flag flapping in the wind or a skirt worn by a character) which allows them to achieve higher detail. Our method starts by pre-computing a pair of coarse and fine training simulations aligned with tracking constraints using harmonic test functions. Next, we train the upsampling operators with a new regularization method that enables us to learn mid-scale details without overfitting. We demonstrate generalizability to unseen conditions such as different
wind velocities or novel character motions. Finally, we discuss how to re-introduce high-frequency details not explainable by the coarse mesh alone using oscillatory modes.
Volume Shadows
Light scattering in a participating medium is responsible for several important effects we see in the natural world. In the presence of occluders, computing single scattering requires integrating the illumination scattered towards the eye along the camera ray, modulated by the visibility towards the light at each point. Unfortunately, incorporating volumetric shadows into this integral, while maintaining real-time performance, remains challenging.
In this paper we present a new real-time algorithm for computing volumetric shadows in single-scattering media on the GPU. This computation requires
evaluating the scattering integral over the intersections of camera rays with the shadow map, expressed as a 2D height field. We observe that by applying epipolar rectification to the shadow map, each camera ray only travels through a single row of the shadow map (an epipolar slice), which allows us to find the visible segments by considering only 1D height fields. At the core of our algorithm is the use of an acceleration structure (a 1D min-max mipmap)
which allows us to quickly find the lit segments for all pixels in an epipolar slice in parallel. The simplicity of this data structure and its traversal allows for
efficient implementation using only pixel shaders on the GPU
Volumetric Obscurance
Obscurance and Ambient Occlusion (AO) are popular techniques in both film and games that model how ambient light is shadowed. While it is largely a solved problem for static scenes, for dynamic scenes it is still difficult to compute at interactive rates. Recent attempts to compute AO in screen space for dynamic scenes either have poor performance or suffer from under-sampling problems. We formulate the problem as a 3D volumetric integral, which maps more naturally to graphics hardware. This integral can be solved using line samples to improve the under-sampling problems that plague other techniques. Following the idea of line integrals to its logical conclusion, we show results using area samples that use a simple statistical model of the depth buffer that allows us to use a single sample. We also discuss strategies for generating point, line, and area sample patterns along with ways to incorporate the surface normal into the volume obscurance calculation.
