Abstract: A method and system provide for augmenting a photograph. An unlabeled photograph is obtained. A weakly augmented photograph and a strongly augmented photograph are obtained from the unlabeled photograph based on different types of data augmentation methods. The weakly augmented photograph is processed through a model to generate multiple weakly augmented photograph class predictions (with assigned probabilities). The multiple weakly augmented photograph class predictions are converted into positive pseudo-labels (indicating a presence of a class) or negative pseudo-labels (indicating absence of a class) using different fixed percentile thresholds. The strongly augmented photograph is processed through the model to generate a strongly augmented photograph class prediction. The model is trained to make the strongly augmented photograph label prediction match the positive pseudo-label via a cross-entropy loss. The trained model is then utilized to label the unlabeled photograph with multiple labels.

Abstract: In various embodiments, a training application trains machine learning models to perform tasks associated with 3D CAD objects that are represented using B-reps. In operation, the training application computes a preliminary result via a machine learning model based on a representation of a 3D CAD object that includes a graph and multiple 2D UV-grids. Based on the preliminary result, the training application performs one or more operations to determine that the machine learning model has not been trained to perform a first task. The training application updates at least one parameter of a graph neural network included in the machine learning model based on the preliminary result to generate a modified machine learning model. The training application performs one or more operations to determine that the modified machine learning model has been trained to perform the first task.

Abstract: A computer-implemented method for generating a virtual collection of digital materials includes: generating a virtual three-dimensional (3D) design workspace; and, in response to a first operation that is associated with a first media file: automatically generating a virtual 3D representation of the first media file; and automatically incorporating the virtual 3D representation of the first media file into the virtual 3D design workspace.

Abstract: In various embodiments, a style comparison application compares geometric styles of different three dimensional (3D) computer-aided design (CAD) objects. In operation, the style comparison application executes a trained neural network one or more times to map 3D CAD objects to feature map sets. The style comparison application computes a first set of style signals based on a first feature set included in the feature map sets. The style comparison application computes a second set of style signals based on a second feature set included in the feature map sets. Based on the first set of style signals and the second set of style signals, the style comparison application determines a value for a style comparison metric. The value for the style comparison metric quantifies a similarity or a dissimilarity in geometric style between a first 3D CAD object and a second 3D CAD object.

Abstract: Techniques are disclosed for posing three-dimensional (3D) objects in extended reality (XR) environments using multiple hands. In some embodiments, an animation application selects a posing mode based on user interactions and adjusts joints and/or bones associated with a 3D object based on the posing mode and the user interactions. The posing mode can be a symmetric bimanual posing mode in which two joints are selected and adjusted simultaneously, an asymmetric bimanual posing mode a reference joint is selected with one hand to serve as a pivot while a second joint is adjusted by another hand, a virtual hand posing mode in which virtual hands are created to fix joints in place, or a bone manipulation mode in which bones can be selected and adjusted, with adjustments to the bones influencing joints that are connected to the bones.

Abstract: A method and system provide for operating a three-dimensional (3D) computer animation and visual effects application (3D application). Execution of a multi-step 3D animation, modeling, or visual effects operation is initialized. Progression of the operation is controlled via a node-based state machine having a plurality of stage nodes daisy-chained via defined dependencies, each stage node including a condition attribute. Upon activation of a first stage node, a script associated with the first stage node that configures behavior of the 3D application is activated. Application events are monitored. A determination is made that the condition attribute of the first stage node has been satisfied. Upon satisfaction of the condition attribute, execution transitions to a subsequent stage node and a corresponding script is executed that modifies a scene state or animation state.

Abstract: Techniques are disclosed for controlling robotic systems to perform assembly tasks. In some embodiments, a robot control application receives sensor data associated with one or more parts. The robot control application applies a grasp perception model to predict one or more grasp proposals indicating regions of the one or more parts that a robotic system can grasp. The robot control application causes the robotic system to grasp one of the parts based on a corresponding grasp proposal. If the pose of the grasped part needs to be changed in order to assemble the part with one or more other parts, the robot control application determines movements of the robotic system required to re-grasp the part in a different pose. In addition, the robot control application determines movements of the robot system for assembling the part with the one or more other parts based on results of a motion planning technique.

Abstract: A method and system provide for automating drawing. A drawing of two or more entities is obtained and a resolution is determined. Based on the resolution, the drawing is quantized into a cell map in which spatial information is lost. The cell map is a collection of multiple cells stored in a contiguous memory. Each of the multiple cells is quantized geometry data and domain specific information. The cell map is utilized to automate a drawing process workflow faster and more efficiently than relying on conventional geometry data structures.

Abstract: A method, apparatus, and system provide the ability to perform a drop test using a response surface. Inputs including a target safety factor (T-SF), a drop height, and a computer model, are acquired. An initial template is evaluated by computing a safety factor for a set of orientations, for each model point MP in the computer model across a set of times (t). A minimum safety factor is determined and a response surface for the model is generated. An actual safety factor SFm is generated by conducting a drop test simulation of the model based on a point having the minimum safety factor and a corresponding orientation. Consistency/validity of the actual safety factor is compared to the target safety factor and the model either passes or additional points may be added to the set of points and the process repeats based on an updated response surface.

Abstract: A method and system provide for computer aided drawing (CAD) automation. A CAD drawing in a first format that is native to a CAD application is obtained. Markup for the CAD drawing that is in a second non-native format is acquired. A machine learning engine processes the markup using a machine learning model, to identify a markup type, identify markup properties, correlate, the markup to a CAD object needed to update the CAD drawing, and determine, CAD object properties and CAD application properties. The CAD object with the CAD object properties is provided in the CAD application using the CAD application properties. User input is accepted and the CAD drawing is updated based thereon. The machine learning model is updated based on the user input and utilized to process future markups.

Abstract: A method and system provide the ability to generate and use synthetic data to extract elements from a floor plan drawing. A room layout is generated. Room descriptions are used to generate and place synthetic instances of symbol elements in each room. A floor plan drawing is obtained and pre-processed to determine a drawing area. Based on the synthetic data symbols in the floor plan drawing are detected. Based on the detected symbols, building information model (BIM) elements are fetched and placed in the floor plan drawing.

Abstract: A computer-implemented method and system provide for insights/recommendations to a user of a software product. A computer application autonomously determines a count of distinct commands executed by a user. Based on the count, the computer application autonomously determines the insight that includes a feature or new distinct command. The computer application then autonomously recommends the insight to the user.

Abstract: A technique for generating designs includes: causing one or more candidate designs to be displayed within a virtual-reality (VR) environment; receiving a user input associated with a first candidate design included in the one or more candidate designs via the VR environment; generating a modified design based at least on the user input and the first candidate design; and generating a plurality of output designs via a generative design process based on the modified design.

Abstract: Methods, systems, and apparatus, including medium-encoded computer program products, for computer aided design of physical structures include: obtaining a B-Rep model including parametric surfaces and associated coedges; generating feature matrices from the coedges associated with the parametric surfaces; using the feature matrices in a convolution layer of a convolutional neural network of a machine learning algorithm to recognize a collection of faces or edges in the B-Rep model, including concatenating results of transforming the feature matrices using a topological walk matrix that represents topological information for the parametric surfaces based on adjacency relationships among the coedges (e.g.

Abstract: Methods, systems, and apparatus, including medium-encoded computer program products, for computer aided design of physical structures include: displaying, in an extended reality (XR) environment of an interconnected cross-review collaboration application, a shared three-dimensional (3D) model in a first spatial review mode; receiving, in the XR environment of the interconnected cross-review collaboration application and by a first user, an indication specifying a first location within the shared 3D model, the indication being defined using either 2D data or 3D data associated with the 3D model; and displaying, in the XR environment of the interconnected cross-review collaboration application, a portion of the 3D model in a second one-to-one scale review mode, the portion of the 3D model being displayed from a perspective associated with the specified first location, wherein the first spatial review mode and the second one-to-one scale review mode are interconnected.

Abstract: One embodiment of the present application sets forth a method for playback of a sense-making operation. The method includes receiving first session data that includes a set of timeline steps. Each timeline step included in the set of timeline steps corresponds to a user action performed on a data set. The method further includes receiving a playback command to display a first sequence of timeline steps included in the set of timeline steps. The method further includes rendering a first graph for display based on at least one timeline step included in the first sequence of timeline steps.

Abstract: Methods, systems, and apparatus, including medium-encoded computer program products, for computer aided design of physical structures using generative design processes, where the three dimensional (3D) models of the physical structures are produced in accordance with a design criterion that limits a minimum thickness of the generatively designed 3D models, include: obtaining a design space for an object to be manufactured and one or more design criteria including a thickness constraint; iteratively modifying a generatively designed 3D shape of the modeled object in the design space in accordance with the one or more design criteria, including measuring a current thickness for the 3D shape using an overall relationship of a volume of the 3D shape with respect to a surface area of the 3D shape; and providing the generatively designed model for use in manufacturing the physical structure using one or more computer-controlled manufacturing systems.

Abstract: Methods, systems, and apparatus, including medium-encoded computer program products, for computer aided design with feature thickness control, include: a three-dimensional modeling program configured to provide voxelized thinning including preparing a voxelized sheet and line skeleton for a three-dimensional shape of a three-dimensional model, and/or ramped scaling based thickness constraint application during shape and/or topology generation including modifying an amount of change indicated by a current numerical assessment generated by numerical simulation. The three-dimensional modeling program can be an architecture, engineering and/or construction program (e.g., building information management program), a product design and/or manufacturing program (e.g., a CAM program), and/or a media and/or entertainment production program (e.g., an animation production program).

Abstract: In various embodiments, a computer animation application automatically solves inverse kinematic problems when generating object animations. The computer animation application determines a target vector based on a target value for a joint parameter associated with a joint chain and at least one of a target position or a target orientation for an end-effector associated with the joint chain. The computer animation application executes a trained machine learning model on the target vector to generate a predicted vector that includes data associated with multiple joint parameters associated with the joint chain.

Abstract: A method and system provide the ability to align a railway track. Survey representing the track are obtained. A centerline of the railway track and curvatures are autonomously computed based on the survey points. Geometry is detected based on the curvature plot and provides for geometry types including a tangent, a spiral, and a curve. An alignment is displayed and includes a curvature plot that includes the detected geometry with each geometry type displayed in a visually distinguishable manner. Parameters of the detected geometry are changed and the alignment is dynamically updated. The alignment is then exported to a physical machine that aligns the railway track consistent with the alignment.