公开(公告)号：US20250077874A1

公开(公告)日：2025-03-06

申请号：US18457854

申请日：2023-08-29

Applicant: Autodesk, Inc.

Inventor： Kaveh Hassani ,  Hyunmin Cheong ,  Adam Noble Arnold ,  Kamal Rahimi Malekshan

IPC: G06N3/088 ,  G06F18/2321 ,  G06F30/12

Abstract: A method and system provide the ability to utilize three-dimensional (3D) models to perform a predictive task. Multiple 3D models, consisting of non-Euclidean data, are obtained. Each 3D model is translated into a relational graph with nodes and edges. Each relational graph is processed using a graph neural network (GNN) that computes a node representation per node. The node representations are aggregated into a structural representation of the 3D model. Multiple different views of the 3D model are captured and passed through a convolutional neural network (CNN) to compute a view representation of each view. The view representations are aggregated into a single visual representation. The GNN and CNN are trained using a multiview contrastive training objective to maximize agreement between the structural representation and the single visual representation to form final learned representations. The final learned representation is utilized to perform the predictive task.

公开(公告)号：US11586464B2

公开(公告)日：2023-02-21

申请号：US16705133

申请日：2019-12-05

Applicant: AUTODESK, INC.

Inventor： Tovi Grossman ,  George Fitzmaurice ,  Benjamin Lafreniere ,  Juho Kim ,  Minsuk Chang ,  Kaveh Hassani

IPC: G06F9/48 ,  G06Q10/06 ,  G06N20/00 ,  G06F30/12 ,  G06F16/901 ,  G06Q10/0633

Abstract: A W-graph system comprising a server connected with a plurality of clients via a network. Each client/user performs a design task via a design application while the server collects timestamped event data. The server generates a plurality of W-graphs for a plurality of tasks based on the collected event data. Each W-graph comprises one or more representative workflows, each representative workflow comprising at least one merged node representing nodes from different workflows for different users performing the same task. A W-graph for a task selected by the user may be viewed in a W-graph GUI. A user may also select a W-suggest function to have a current workflow for a task analyzed for optimization based on a W-graph generated for the same task. A modified current workflow is generated that highlights user techniques in the current workflow that are less efficient than user techniques in the W-graph.