Voice driven 3D static asset creation in computer simulations

1 . A non-transitory, computer readable storage medium storing instructions that, when executed by a computing system, cause the computing system to perform operations comprising:

receiving text;

executing or requesting execution of an artificial intelligence (AI) engine, wherein, the text is a first input to the AI engine, and wherein the AI engine outputs, based on the text, a two-dimensional (2D) image of a computer simulation asset;

converting, the 2D image to a three-dimensional (3D) asset;

receiving a second input indicating a change to the 3D asset;

modifying an encoding for the 3D asset based on the second input to generate a 3D model; and

presenting, the 3D model within a computer simulation environment that includes another computer simulation asset.

2 . The non-transitory, computer readable storage medium of claim 1 , wherein the text is extracted from voice-to-text conversion or direct text input.

3 . The non-transitory, computer readable storage medium of claim 1 , wherein the operations further comprise:

converting voice input to the text;

extracting asset descriptions or location; and

associating the asset description or the location with the 3D asset based at least in part on the text.

4 . The non-transitory, computer readable storage medium of claim 1 , wherein the text comprises at least one location, such that, the 3D model is presented to the computer simulation environment at the location relative to other computer simulation assets.

5 . The non-transitory, computer readable storage medium of claim 2 , wherein the voice-to-text conversion indicates at least plural objects and the 3D model is consistent with the plural objects.

6 . The non-transitory, computer readable storage medium of claim 1 , the operations further comprising:

modifying the encoding of the 3D asset at least in part by varying weights of interpolated pixels along at least one border region in the 3D asset;

generating the 3D model of the 3D asset from the encoding; and

presenting the 3D model to the computer simulation environment that includes another computer simulation asset.

7 . The non-transitory, computer readable storage medium of claim 6 , the operations further comprising:

receiving the second input to modify the encoding of the 3D asset.

8 . A method comprising:

receiving a first input, wherein the input comprises at least an image or text, corresponding to a two-dimensional (2D) object;

generating a 2D synthetic representation of the 2D object;

converting the 2D synthetic representation to a 3D asset;

receiving a second input indicating a change to the 3D asset;

modifying an encoding for the 3D asset based on the second input to generate a 3D model; and

presenting the 3D model within a computer simulation environment that includes another computer simulation asset.

9 . The method of claim 8 , wherein generating the 2D synthetic representation includes:

executing or requesting execution of an artificial intelligence (AI) engine;

inputting the text or image into the AI engine, wherein the AI engine comprises at least one conditional generative neural model or a generative adversarial network (GAN); and

performing at least one of interpolation and feature editing using a 2D transformation model to generate the 2D synthetic representation.

10 . The method of claim 8 , further comprising:

converting the first input to indicate at least one location; and

presenting the 3D model within the computer simulation environment at the at least one location relative to other computer simulation assets.

11 . The method of claim 8 , further comprising:

receiving speech that indicates at least plural objects; and

assessing whether the 3D model is consistent with the plural objects.

12 . The method of claim 8 , further comprising:

transmitting the 3D asset to a user for modification;

receiving the second input from the user indicating a change to the 3D asset; and

generating the 3D model based on the modified encoding of the 3D asset.

13 . The method of claim 8 , wherein receiving the image or text includes providing a user interface (UI) including a selector for use in uploading the image or text.

14 . The method of claim 8 , further comprising receiving a description of a desired asset scene.

15 . An apparatus comprising:

at least one processor;

at least one computer output device controlled by the at least one processor; and

a non-transitory computer readable medium storing instructions that, when executed by the at least one processor, cause the at least one processor to perform operations comprising:

identifying a two-dimensional (2D) image;

generating a synthetic representation of the 2D image via an artificial intelligence (AI) engine;

converting the synthetic representation of the 2D image to a reconstructed mesh of a 3D asset, wherein the 3D asset comprises at least asset characteristics or location;

receiving an input indicating a change to the 3D asset;

modifying an encoding for the 3D asset based on the input; and

presenting the reconstructed mesh of the 3D asset as an object within a computer simulation environment at the location and relative to other computer simulation assets.

16 . The apparatus of claim 15 , wherein the non-transitory computer readable medium stores additional instructions that when executed by the at least one processor, cause the at least one processor to perform additional operations comprising:

modifying a geometry of the 3D asset using at least one physics engine operating on one or more inertial tensors to define how the 3D asset reacts to a simulated force in the computer simulation environment.

17 . The apparatus of claim 15 , wherein the non-transitory computer readable medium stores additional instructions that when executed by the at least one processor, cause the at least one processor to perform additional operations comprising:

identifying the 2D image based at least in part on text.

18 . The apparatus of claim 17 , wherein the non-transitory computer readable medium stores additional instructions that when executed by the at least one processor, cause the at least one processor to perform additional operations comprising:

extracting the text from an image or voice input.

19 . The apparatus of claim 17 , wherein the non-transitory computer readable medium stores additional instructions that when executed by the at least one processor, cause the at least one processor to perform additional operations comprising:

inputting the text into the AI engine to generate the 2D image based at least in part on the text.

20 . The apparatus of claim 15 , wherein the non-transitory computer readable medium stores additional instructions that when executed by the at least one processor, cause the at least one processor to perform additional operations comprising:

generating, based on at least one of geometry or texture information of the 3D asset, encodings for the 3D asset using different weights;

using the encodings to train a classifier to discover at least one feature axis; and

using the feature axis to modify the 3D asset to render a modified asset.

21 . The apparatus of claim 20 , wherein the non-transitory computer readable medium stores additional instructions that when executed by the at least one processor, cause the at least one processor to perform additional operations comprising:

rendering the modified asset at least in part by modifying the encoding of the 3D asset using a product of a variable and the at least one feature axis.

22 . The apparatus of claim 15 , wherein the non-transitory computer readable medium stores additional instructions that when executed by the at least one processor, cause the at least one processor to perform additional operations comprising:

modifying the 3D asset based at least in part on physics modeling of environmental impact on the 3D asset.