Pixar

Seriforge Inc.

Apple Inc.

Pixar Animation Studio

Eric Gregory

Oren M Jacob

Robert Jensen

Michael S Ferris

Timothy S Milliron

Jonathan Worthy Hollander

Brett Levin

Brad Andalman

Mike Ferris

Thomas Benedict Gunter

Wilson Taylor Holliday

Marco Zvanik

Andrew Justin Butts

Keith Scott Brisson

Thomas Robert Nickson

Arthur A Van Hoff

Systems and processes for creating and updating natural language models are provided. An example process of creating a natural language model includes, at an electronic device with one or more processors and memory, receiving an utterance, associating an action structure with the utterance, determining a plurality of augmented utterances based on the received utterance, creating a natural language model including the received utterance and the plurality of augmented utterances by mapping the plurality of augmented utterance to the associated action structure, and providing the natural language model including the received utterance and the plurality of augmented utterances to a second electronic device.

Using augmentation to create natural language models

Various embodiments provide an apparatus, method, and/or system by which fiber, e.g., in the form of tow or tow, is used to create a shaped ply by facilitating drawing fiber from a supply along a fiber axis between two sets of pins; moving the pins across the fiber axis to form a fiber web in the desired ply shape; fixing the fiber web to form the shaped ply; and releasing the shaped ply. By enabling formation of shaped plies (including plies with doubly curved surfaces and simultaneous sequences of strategically shaped plies) directly from a single piece of fiber tow, various embodiments discussed herein enable substantial reductions in labor and materials costs that are conventionally associated with construction of composite preforms and accompanying composite products.

Systems and methods of converting fiber into shaped fabric plies for composite preforms and products

A process to create composite preforms and parts by stacking layers of two-dimensional fiber material (e.g., fiber cloth, fiber reinforced fabric, etc.) having pin-receiving holes and/or gaps disposed between the fibers of the fiber material is described. The stack is pinned together using a subset of pins, such as fiber reinforcing pins, inserted into a subset of the pin receiving holes and/or gaps, leaving at least some of the pin receiving holes and/or gaps available for pinning other layers, as layers of fiber material are added to the stack. As additional layers are added to the stack, different subsets of pins connect the additional layers to the stack, thereby building up the stack. Each layer of fiber material may have a different shape than the other layers and any arbitrary topology, potentially including non-convex and/or disjoint shapes. Furthermore, implementations may produce composite preforms having an arbitrary number of interconnected layers.

System and method for pinning fiber material

Systems and methods for customizing animation variables and modifications to animation variables in an animation system are provided. An animated model may be comprised of a hierarchical structure of rigs and sub-rigs. An animator may customize the location of animation variables within the hierarchical structure through a relocation operation from an original position to a relocated position. The animation system identifies the relocation operation, resulting in an association being generated between the original position and the relocated position. Modifications made to animation variables in the animation system may be received by the animation system and the animator can customize the scope of the modification and its application to the animated model or animated scene.

File path translation for animation variables in an animation system

A method allowing for quick manipulation of weight values associated with points on a polygonal mesh that is to be deformed. A point on the polygonal mesh may be selected by the user. Then, a solution space of possible positions for the selected point may be calculated by solving a deformation model for a range of weight values. A graphical representation of the solution space may be provided, such as a locus of possible positions for the selected point, where each point on the locus corresponds to a particular weight value. Manipulation of these weight values, and hence, the deformation of the polygonal mesh, may be achieved simply by selecting a position on the locus. The mesh may be updated to reflect the weight corresponding to the selected position.

Graphical user interface for performing deformations

A user interface presents views of entities and their attributes relative to any subset of associated parameters. A view may present views of entities as a function of one or more parameters. A selection of a subset of entities in one view with respect to parameters may be presented in another view with respect to different parameters. Entities may be manipulated in any supplemental view to change their values or their parameter associations. A view may present components and parameters associated with layers or any subset of layers and/or after layer compositing. A view can include resolved values of components and their attributes. A supplemental view may include visual indicators to indicate the layer or other parameter defining its resolved value. Hierarchical relationships may be presented in views, with ancestor entities expanded within a view to show its descendent entities or collapsed to hide its descendent entities.

Visualization of value resolution for multidimensional parameterized data

Multiple time scales enable animations over multiple shots. Time-dependent data is specified with reference to time references. Data expressed in different time references may be combined to form animations, shots, and sequences of shots. Time transformations specifying mappings between time references are used to combine data expressed in different time references. Data can be viewed in arbitrary time references. Editing data specifying a sequence of shots may be used as a time reference for viewing data expressed in other time references. Data expressed in one time reference may be rearranged according to another time reference to indicate time reference relationships. Data authored in one time reference may be viewed and modified in another time reference. Modifications are converted from the viewing time reference to the data's original time reference for updating data. Time-dependent data may be expressed in layer data structures. Layer relationships may include time transformations.

Multiple time scales in computer graphics

Components are presented to users based on the components' assigned task memberships. The tasks represents the different activities performed by users within a digital production pipeline. Each stage of the digital production pipeline may be associated with a task. Additional user-defined tasks may be defined as needed to further distinguish components. Attributes of components are associated with one or more tasks when the component is authored. Task memberships may be inherited from parent components or overridden. A user interface includes a task selection menu to receive a selection of one or more tasks from the user. The task selection is used to filter components, so that the user interface presents only relevant information to users. Component attribute values and task associations can be expressed using layer data structures. Task restrictions of layers allow layers to only specify values of components that match the task associations of the layer.

Task membership and task masks

The components of a computer graphics scene can be organized using multiple independent graphs. Each component can include graph relationship attributes specifying its location in one or more graphs. Graph can be utilized for any purpose within the digital production pipeline. A component can include any number of graph relationship attributes. The number and type of graph relationship attributes can be different for each component. Software tools can access one or more graphs to determine structural or functional relationships between components. Software tools can enforce policies for a graph independent of other graphs. Software tools can optimize the components of a graph independent of other graphs. A graph relationship attribute of a component can include optional parameter data to specify properties of the relationship between two or more components. Layer data structures including opinions of value can specify graph relationships.

Graphs of components for digital productions

A unified mechanism of scene description includes layering and cross-referencing relationships. The scene description specifies components using one or more layers. A layer includes an opinion of value for one or more attributes of one or more components. A layer can be sparsely populated with opinions of attribute values for only portions of components' attributes. Layers express opinions of value for an attribute using the semantic structure associated with the attribute. The scene description organizes layers in a layer hierarchy, which determines a layer stacking order. The layer stacking order specifies the compositing of opinions of attribute values to determine the complete authored state of the component. The layer stacking order enables attributes of components to be selectively overridden without disturbing other attribute values. Cross-referencing relationships enable encapsulation and reuse of components. Cross-referencing relationships can encapsulate components defined by one or more layers.

Growing community of inventors

Larkspur, CA, United States of America

Eric Gregory