Imputation of 3D data using generative adversarial networks

1. A non-transitory computer readable storage medium having stored thereon instructions that, when executed by one or more processors, cause a computer to:

generate a loss value by processing one or more three-dimensional regions and at least one three-dimensional point cloud;

update one or more weights of a generative adversarial network by backpropagating the loss value;

store the updated weights of the generative adversarial network on a non-transitory computer readable storage medium;

obtain a three-dimensional point cloud having one or more gaps;

initialize the generative adversarial network using the stored weights; and

impute one or both of (i) RGB data, and (ii) elevation data into the gaps of the three-dimensional point cloud by analyzing the three-dimensional point cloud using the initialized generative adversarial network.

2. The non-transitory computer readable storage medium of claim 1 , having stored thereon further instructions that, when executed by one or more processors, cause a computer to:

store the three-dimensional point cloud including the imputed data on the computer readable storage medium.

3. The non-transitory computer readable storage medium of claim 2 , wherein the gaps include one or both of (i) an implicit gap, and (ii) an explicit gap.

4. The non-transitory computer readable storage medium of claim 1 , having stored thereon further instructions that, when executed by one or more processors, cause a computer to:

generate the three-dimensional point cloud using a structure-from-motion technique.

5. The non-transitory computer readable storage medium of claim 1 , having stored thereon further instructions that, when executed by one or more processors, cause a computer to:

backpropagate discriminator loss to a discriminator artificial neural network.

6. The non-transitory computer readable storage medium of claim 1 , having stored thereon further instructions that, when executed by one or more processors, cause a computer to:

backpropagate discriminator loss to a discriminator artificial neural network and a generator artificial neural network.

7. A computer-implemented method for training a generative adversarial network, comprising:

generating a loss value by processing one or more three-dimensional regions and at least one three-dimensional point cloud;

updating one or more weights of the generative adversarial network by backpropagating the loss value;

storing the updated weights of the generative adversarial network on a non-transitory computer readable storage medium;

obtaining a three-dimensional point cloud having one or more gaps;

initializing the generative adversarial network using the stored weights; and

imputing one or both of (i) RGB data, and (ii) elevation data into the gaps of the three-dimensional point cloud by analyzing the three-dimensional point cloud using the initialized generative adversarial network.

8. The computer-implemented method of claim 7 , further comprising:

storing the three-dimensional point cloud including the imputed data on the computer readable storage medium.

9. The computer-implemented method of claim 7 , wherein obtaining the three-dimensional point cloud having the one or more gaps includes generating the three-dimensional point cloud using a structure-from-motion technique.

10. The computer-implemented method of claim 9 , wherein the gaps include one or both of (i) an implicit gap, and (ii) an explicit gap.

11. The computer-implemented method of claim 7 , wherein updating the one or more weights of the generative adversarial network by backpropagating the loss value throughout the generative adversarial network includes backpropagating discriminator loss to a discriminator artificial neural network.

12. The computer-implemented method of claim 7 , wherein updating the one or more weights of the generative adversarial network by backpropagating the loss value throughout the generative adversarial network includes backpropagating discriminator loss to a discriminator artificial neural network and a generator artificial neural network.

13. A computing system for training a generative adversarial network, the system comprising:

one or more processors; and

a memory storing instructions that, when executed by the one or more processors, cause the system to:

generate a loss value by processing one or more three-dimensional regions and at least one three-dimensional point cloud;

update one or more weights of the generative adversarial network by backpropagating the loss value;

store the updated weights of the generative adversarial network on a non-transitory computer readable storage medium;

obtain a three-dimensional point cloud having one or more gaps;

initialize the generative adversarial network using the stored weights; and

impute one or both of (i) RGB data, and (ii) elevation data into the gaps of the three-dimensional point cloud by analyzing the three-dimensional point cloud using the initialized generative adversarial network.

14. The computing system of claim 13 , wherein the gaps include one or both of (i) an implicit gap, and (ii) an explicit gap.

15. The computing system of claim 13 , the memory having stored thereon further instructions that, when executed by the one or more processors, cause the computing system to

backpropagate discriminator loss to a discriminator artificial neural network.

16. The computing system of claim 13 , the memory having stored thereon further instructions that, when executed by the one or more processors, cause the computing system to

backpropagate discriminator loss to a discriminator artificial neural network and a generator artificial neural network.

17. The computing system of claim 13 , the memory having stored thereon further instructions that, when executed by the one or more processors, cause the computing system to

display the three-dimensional point cloud including the imputed data in a display device of a user.