Dassault Systemes has been granted a patent for a computer-implemented method that trains a neural network to infer a solid CAD feature from a freehand drawing representing a 3D shape. The method utilizes a dataset of freehand drawings to improve the accuracy of the inference process. GlobalData’s report on Dassault Systemes gives a 360-degree view of the company including its patenting strategy. Buy the report here.

According to GlobalData’s company profile on Dassault Systemes, was a key innovation area identified from patents. Dassault Systemes's grant share as of February 2024 was 58%. Grant share is based on the ratio of number of grants to total number of patents.

Training neural network to infer 3d shapes from freehand drawings

Source: United States Patent and Trademark Office (USPTO). Credit: Dassault Systemes SE

A recently granted patent (Publication Number: US11922573B2) discloses a computer-implemented method for training a neural network to infer solid CAD features from freehand drawings representing 3D shapes. The method involves obtaining a dataset of freehand drawings, training the neural network based on this dataset, and configuring the network to interpret unparameterized 2D image data structures directly input by users without transformation. The neural network utilizes recurrent neural network (RNN) cells to infer curves of the solid CAD feature and includes candidate sets of curves and sweep operations for further training.

Furthermore, the patent describes a device comprising a processor and a data storage medium with a recorded data structure representing the neural network. The device is capable of obtaining freehand drawings, applying the neural network to infer solid CAD features, and includes instructions for training the network based on the dataset. The processor is configured to handle candidate sets of curves and sweep operations, utilizing RNN cells for inference. The method also includes supervised training to minimize loss and unsupervised training to penalize disparities between freehand drawings and inferable CAD features, enhancing the network's accuracy and performance in interpreting 3D shapes from user sketches.

To know more about GlobalData’s detailed insights on Dassault Systemes, buy the report here.

Premium Insights


The gold standard of business intelligence.

Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.


GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.

GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.