Generating flower images and shapes with compositional pattern producing networks
Various embodiments are disclosed for generating an image from a Compositional Pattern Producing Network (CPPN). One such method includes receiving, in the CPPN, a series of polar coordinates {r,θ}; outputting, by the CPPN, a series of pixel values, each of the pixel values corresponding to one of the polar coordinates; and displaying the pixel values at the corresponding polar coordinates to produce the image.
1. A method comprising:
encoding an image in a Compositional Pattern Producing Network (CPPN), the CPPN including a plurality of connected nodes, the CPPN receiving as an input a series of polar coordinates {r,θ} and producing as an output a series of pixel values;
querying the CPPN for one of the pixel values at each of the polar coordinates; and
displaying the series of pixel values at the corresponding polar coordinates to produce the image.
2. The method of claim 1 , wherein the pixel values are expressed as gray scale values.
3. The method of claim 1 , wherein the pixel values are expressed as color values in a color space.
4. The method of claim 3 , wherein the color space is defined by hue, saturation, and value (HSV).
5. The method of claim 3 , wherein the color space is defined by red, green, and blue (RGB).
6. The method of claim 1 , wherein the image is a three-dimensional image.
7. The method of claim 1 , wherein the CPPN also receives a layer input L, and wherein the querying further comprises querying the CPPN at a plurality of layers 1 to L.
8. The method of claim 7 , wherein the image is a two-dimensional image.
9. A method of generating an image from a Compositional Pattern Producing Network (CPPN), the CPPN comprising a plurality of nodes connected in a graph, each of the nodes associated with an activation function, the method comprising:
receiving, in the CPPN, a series of polar coordinates {r,θ};
outputting, by the CPPN, a series of pixel values, each of the pixel values corresponding to one of the polar coordinates; and
displaying the pixel values at the corresponding polar coordinates to produce the image.
10. The method of claim 9 , wherein the pixel values are expressed as gray scale values.
11. The method of claim 9 , wherein the pixel values are expressed as color values in a color space.
12. The method of claim 9 , wherein the displaying comprises translating the corresponding polar coordinates into cartesian coordinates.
13. The method of claim 9 , further comprising translating the corresponding polar coordinates into cartesian coordinates, and wherein the pixel values are displayed at the cartesian coordinates.
14. A method of generating an image from a Compositional Pattern Producing Network (CPPN), the CPPN comprising a plurality of nodes connected in a graph, each of the nodes associated with an activation function, the method comprising:
receiving, in the CPPN, a series of polar coordinates, each of the polar coordinates including a radius coordinate r and an angular coordinate θ; and
applying a periodic transform P to each of the angular coordinates θ to produce a transformed polar coordinate;
outputting, by the CPPN, a series of pixel values, each of the pixel values corresponding to one of the transformed polar coordinates; and
displaying the pixel values at the corresponding transformed polar coordinates to produce the image.
15. The method of claim 14 , wherein the pixel values are expressed as gray scale values.
16. The method of claim 14 , wherein the pixel values are expressed as color values in a color space.
17. The method of claim 14 , wherein the periodic transform P is a trigonometric function.
18. The method of claim 14 , wherein the periodic transform P is a sine function.
19. A method comprising:
encoding an image in a Compositional Pattern Producing Network (CPPN), the CPPN including a plurality of connected nodes, the CPPN receiving as input a parameter L and a series of polar coordinates and producing as output a series of pixel values, L representing a number of layers;
for a first one of the layers, querying the CPPN for one of the pixel values at each of the polar coordinates and for a first value of L, and displaying the pixel values at the corresponding polar coordinates to produce the image; and
for the remaining ones of the layers, querying the CPPN for one of the pixel values at each of the polar coordinates and for successive values of L, and displaying the pixel values at a scaled version of each the corresponding coordinates to produce the image.
20. The method of claim 19 , wherein the image is a two-dimensional image.