US 11,380,045 B2 · Shaped-based techniques for exploring design spaces

Abstract

In various embodiments, a training application generates a trained encoder that automatically generates shape embeddings having a first size and representing three-dimensional (3D) geometry shapes. First, the training application generates a different view activation for each of multiple views associated with a first 3D geometry based on a first convolutional neural network (CNN) block. The training application then aggregates the view activations to generate a tiled activation. Subsequently, the training application generates a first shape embedding having the first size based on the tiled activation and a second CNN block. The training application then generates multiple re-constructed views based on the first shape embedding. The training application performs training operation(s) on at least one of the first CNN block and the second CNN block based on the views and the re-constructed views to generate the trained encoder.

Claims20

Dependent claims

A computer-implemented method for generating computational representations for three-dimensional (3D) geometry shapes, the method comprising:

for each view included in a first plurality of views associated with a first 3D geometry, generating a view activation based on a first convolutional neural network (CNN) block;
aggregating the view activations to generate a first tiled activation;
generating a first shape embedding that includes a first vector having a fixed size based on the first tiled activation and a second CNN block, wherein the fixed size is selected to control a maximum amount of compute resources for comparing the first shape embedding with a shape embedding generated for another 3D geometry;
generating a first plurality of re-constructed views based on the first shape embedding;
performing one or more training operations on at least one of the first CNN block or the second CNN block based on the first plurality of views and the first plurality of re-constructed views to generate a trained encoder;
and
generating a second shape embedding that includes a second vector having the fixed size based on a second 3D geometry and the trained encoder.

4 more elements

One or more non-transitory computer readable media including instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of:

for each view included in a first plurality of views associated with a first 3D geometry, generating a view activation based on a first convolutional neural network (CNN) block;
aggregating the view activations to generate a first tiled activation;
generating a first shape embedding that includes a first vector having a fixed size based on the first tiled activation and a second CNN block, wherein the fixed size is selected to control a maximum amount of compute resources for comparing the first shape embedding with a shape embedding generated for another 3D geometry;
generating a first plurality of re-constructed views based on the first shape embedding;
performing one or more training operations on at least one of the first CNN block or the second CNN block based on the first plurality of views and the first plurality of re-constructed views to generate a trained encoder;
and
generating a second shape embedding that includes a second vector having the fixed size based on a second 3D geometry and the trained encoder.

4 more elements

Dependencies are inferred from the claim text.

Shaped-based techniques for exploring design spaces

Life of this patent

Shaped-based techniques for exploring design spaces

Life of this patent

Abstract

Claims20