Abstract: A scene generation system provide graphics enhanced content placement in the quality or fidelity of promoted visual assets in a scene are adjusted relative to other or non-promoted visual assets in the same scene. The different quality or fidelity at which the promoted visual assets are presented draw viewer focus to the promoted visual assets without prominently positioning the promoted visual assets in the scene or otherwise altering the creative intent for the scene. The system receives a request to promote a particular visual asset in a scene, determines a first level-of-detail (LoD) at which to render the visual assets of the scene, selects a second LoD that is greater than the first LoD for the particular visual asset, and presents the scene by rendering the non-promoted visual assets at the first LoD and the particular visual asset at the second LoD.

Abstract: A splat generation system and associated methods implement a modified splatting pipeline that quantifies loss at each splat generation iteration without rendering the splats generated at each iteration or comparing the resulting visualization against reference images of a 3D asset. The system receives the 3D asset and generates a splat representation of the 3D asset. The system compares geometric similarity between different sets of generated splats in different regions of the splat representation and different sets of 3D primitives in corresponding regions of the 3D asset. The system adjusts a set of splats in response to the geometric similarity between the set of splats and a set of 3D primitives in corresponding regions not satisfying a threshold, and reconstructs the 3D asset with an acceptable amount of loss based on the adjustments made to the splats on a region-by-region basis.

Abstract: A three-dimensional (3D) animation system and associated methods generate a musculoskeletal framework for a 3D model, automatically rig virtual muscles and virtual bones of the musculoskeletal framework to the 3D model, and realistically animate the 3D model based on real musculoskeletal movements associated with the virtual muscles and the virtual bones. The 3D animation system receives multiple scans of a subject, generates primitives that form a 3D model of the subject based on a first scan, and rig the 3D model for animation with the virtual muscles and virtual bones of the musculoskeletal framework that are defined from data of other scans. The 3D animation system animates the 3D model by determining an association between a virtual muscle and a set of primitives created from the rigging, and by adjusting the set of primitives according to a movement created from a simulated contraction of the virtual muscle.

Abstract: Disclosed is a system and associated methods for color correcting three-dimensional (“3D”) objects by leveraging the 3D positional data associated with point cloud data points that form the 3D objects and by adjusting the defined or inherited color values of the data points to account for multiple factors that are derived from the data point 3D positions. The system determines a color variance between the color values of the first set of points and the color values of the second set of points, determines a variance factor based on the positional elements of the second set of points that contributes to the color variance between the points, and adjusts the color variance to a modified color variance based on the variance factor. The system then modifies or color corrects the color values of the second set of points based on the modified color variance.

Abstract: A splat generation system and associated methods implement splatting with adaptive density control to generate a splat representation of a three-dimensional (3D) model in which the splat density and quality in regions of the 3D model that are prioritized based on creator intent or that are commonly or consistently in a user’s field-of-view are increased relative to the splat density and quality in other regions of the 3D model. The system receives the different priorities associated with different parts of the 3D model based on the creator intent or tracked view paths. The system generates a first set of splats that represent the 3D model with a first fidelity, and generates a second set of splats at a greater second fidelity for a first subset of parts of the 3D model that have a higher priority than a second subset of parts of the 3D model.

Abstract: A streaming system performs a dynamic densification of streamed content based on tracked user focus. The streaming system streams different content having a common classification to one or more users, and tracks a focus of the one or more users on different parts of the different content. The streaming system receives a request for new content, classifies the new content with the common classification, and streams a first parts of new content with greater detail than second parts of the new content in response to the request based on corresponding first parts from the different parts of the different content receiving more of the focus than corresponding second parts from the different parts of the different content.

Abstract: An adaptive streaming system dynamically streams different parts of a three-dimensional (3D) model at different resolutions to maximize visual detail and quality in response to changing network performance and/or client device rendering performance. The system receives a request to view the 3D model from a particular field-of-view. The system associates different priorities to different parts of the 3D model, selects first Gaussian splats associated with nodes at a first level in a tree based on the first Gaussian splats representing the different parts of the 3D model in the particular field-of-view with a first priority, and selects second Gaussian splats associated with nodes at a second level in the tree structure based on the second Gaussians splats representing the different parts of the 3D model in the particular field-of-view with a second priority. The system streams the selected Gaussian splats in response to the request.

Abstract: Systems and associated methods are provided for distributed three-dimensional (3D) content generation whereby different 3D assets for a 3D scene are generated by different asset generators at different network tiers according to the latency sensitivity of each 3D asset. The system retrieves a 3D asset for the requested 3D scene, and differentiates a first 3D asset that is latency sensitive from a second 3D asset that is latency tolerant. The system generates the first 3D asset with a first asset generator at a first network tier, and generates the second 3D asset with a second asset generator at a more distant second network tier. The system distributes the generated primitives for the first 3D asset to the user device with a first amount of latency and the generated primitives for the second 3D asset to the user device with a second amount of latency.

Abstract: An optimization system reduces the data encoded within a point cloud for streaming and/or rendering of a lossy representation of the point cloud. The optimization system generates a first optimized splat for a first visual characteristic of the point cloud by replacing a first set of points having a first common value for the first visual characteristic with a first replacement primitive, and generates a second optimized splat for a second visual characteristic of the point cloud by replacing a second set of points having a second common value for the second visual characteristic with a second replacement primitive. The optimization system provides the first optimized splat and the second optimized splat instead of the original points of the point cloud in response to a request to access the point cloud.

Abstract: A distribution system adaptively provides different lossy encodings for different views of a point cloud to a client device based on network or rendering performance of the client device. The distribution system receives a request to access the point cloud, and determines the one or more performance parameters that limit an amount of the point cloud data that the client device is able to receive or process in a given time. The distribution system selects and provides the client device with different sets of optimized splats for different views of the point cloud that satisfy the one or more performance parameters based on a cumulative amount of data encoded within the different sets of optimized splats being equal to or less than the amount of point cloud data that the client device is able to receive or process in the given time.

Abstract: A three-dimensional (3D) animation system and associated methods generate a musculoskeletal framework for a 3D model, automatically rig virtual muscles and virtual bones of the musculoskeletal framework to the 3D model, and realistically animate the 3D model based on real musculoskeletal movements associated with the virtual muscles and the virtual bones. The 3D animation system receives multiple scans of a subject, generates primitives that form a 3D model of the subject based on a first scan, and rig the 3D model for animation with the virtual muscles and virtual bones of the musculoskeletal framework that are defined from data of other scans. The 3D animation system animates the 3D model by determining an association between a virtual muscle and a set of primitives created from the rigging, and by adjusting the set of primitives according to a movement created from a simulated contraction of the virtual muscle.

Abstract: A streaming system and associated methods provide adaptive streaming of point cloud data for out-of-order presentation of important visual features before less important visual features. The adaptive streaming includes retrieving the points for a requested point cloud, differentiating different sets of points that represent different features in the point cloud, prioritizing each set of points based on a feature that is represented by that set of points, and streaming the different sets of points across a data network to a client device in an order that is determined from the prioritization. Moreover, the adaptive streaming may dynamically select a different subset of points from each set of points and different subset of point data to stream with each selected subset of points to visually convey important detail of each feature without all the corresponding points or data that make up that feature in the point cloud.

Abstract: A splat generation system and associated methods perform splatting based on a controlled sampling of a three-dimensional (3D) asset. The controlled sampling ensures minimal coverage the 3D asset primitives and provides enhanced coverage for primitives that have greater detail or variation. The system projects rays in different directions from a set of the primitives towards a bounding volume, and detects points at which some of the projected rays intersect the bounding volume. The system defines virtual cameras in the 3D space of the 3D asset based on a set of the intersection points, and obtains the minimal and/or enhanced coverage based on one or more views of the primitives captured in images taken by the virtual cameras. The system generates a set of splats from the one or more views in the captured images with an acceptable amount of loss and with less total data than the 3D asset.

Abstract: An editing system and associated methods execute a workflow involving a sequence of two-dimensional/three-dimensional (“2D/3D”) data manipulations and/or associated operations to 2D/3D data of different formats by calling functions of different applications that are compatible with the changing format of the 2D/3D data throughout the workflow. The system determines that the 2D/3D data provided as input to a first workflow node is in a first format, and executes the first node by invoking a function of a first application that implements the operations associated with the first node on the 2D/3D data in the first format. Execution of the first node converts the 2D/3D data to a different second format that is passed to a second workflow node. The system executes a second workflow node by invoking a function of a second application that implements the operations associated with the second node on the 2D/3D data in the second format.

Abstract: A three-dimensional (3D) content creation system modifies operation of a radiance field, neural network, and/or other generative artificial intelligence in order to generate 3D content based on constraints that modify the 3D content modeling. The system receives constraints for reducing a first size of a first 3D representation of a 3D object, and generates different sets of 3D primitives with values for one or more parameters of the 3D primitives that satisfy the constraints. The system selects a particular set of 3D primitives that produces a visual representation that differs from the first 3D representation by less than a threshold amount, and presents the particular set of 3D primitives as a size-optimized second 3D representation of the 3D object with a second size that is less than the first size of the first 3D representation.

Abstract: A real-time streaming system and associated methods are provided for initially presenting three-dimensional (3D) content at a low-fidelity first encoding format so that the initial visualization of the 3D content on the requesting device is presented with no or insignificant delay and for streaming prioritized 3D assets within the initial visualization at higher fidelities in different encoding formats that preserve the real-time responsiveness of the system. The system generates and streams a two-dimensional (2D) image for a first presentation of the 3D content at a first fidelity. The system increases a fidelity of a prioritized 3D asset while the 3D content does not change by streaming 3D primitives for the prioritized 3D asset in a second encoding format that increases the fidelity of the prioritized 3D asset in a second presentation of the 3D content from the first fidelity to a greater second fidelity.

Abstract: A streaming system and associated methods provide adaptive streaming of point cloud data for out-of-order presentation of important visual features before less important visual features. The adaptive streaming includes retrieving the points for a requested point cloud, differentiating different sets of points that represent different features in the point cloud, prioritizing each set of points based on a feature that is represented by that set of points, and streaming the different sets of points across a data network to a client device in an order that is determined from the prioritization. Moreover, the adaptive streaming may dynamically select a different subset of points from each set of points and different subset of point data to stream with each selected subset of points to visually convey important detail of each feature without all the corresponding points or data that make up that feature in the point cloud.

Abstract: A splat generation system and associated methods generate a splat representation of a three-dimensional (3D) model with varying detail based on an unequal or non-uniform capture of the 3D model. The system receives the 3D model. The system analyzes variation in the position or visual characteristics of the 3D model primitives, attributes different levels-of-importance to different surfaces or regions of the 3D model based on the variation, and defines a non-uniform placement of virtual cameras around the 3D model according to the different levels-of-importance that are attributed to the different surfaces or regions of the 3D model. The system generates the splats for the splat representation that reconstruct the different surfaces or regions with differing amounts of detail that match the detail captured for the different surfaces or regions in images generated from the non-uniform placement of the virtual cameras.

Abstract: A multi-dimensional encoder (“MDE”) receives a file with different data points. Each data point is defined with uncompressed multi-dimensional data. The multi-dimensional data may include multiple elements for defining a data point position (e.g., a multi-dimensional position including x, y, and/or z coordinates), color values of the data point (e.g., red, green, and blue), and/or other data point attributes. The MDE assigns an index to each data point or each data point element, and maps the data points to a frequency domain based on a frequency with which values occur in two or more of the elements (e.g., multiple dimensions). The MDE generates a line that represents the frequency, and provides a compressed file format to a requesting device that includes the line and different sets of indices that are associated with different frequencies represented by the line.

Abstract: A streaming system and associated methods provide adaptive streaming of point cloud data for out-of-order presentation of important visual features before less important visual features. The adaptive streaming includes retrieving the points for a requested point cloud, differentiating different sets of points that represent different features in the point cloud, prioritizing each set of points based on a feature that is represented by that set of points, and streaming the different sets of points across a data network to a client device in an order that is determined from the prioritization. Moreover, the adaptive streaming may dynamically select a different subset of points from each set of points and different subset of point data to stream with each selected subset of points to visually convey important detail of each feature without all the corresponding points or data that make up that feature in the point cloud.