Creative GAN generating art deviating from style norms

Patent metadata

Field	Value
Publication number	US-10853986-B2
Application number	US-201916447712-A
Country	US
Kind code	B2
Filing date	Jun 20, 2019
Priority date	Jun 20, 2018
Publication date	Dec 1, 2020
Grant date	Dec 1, 2020

Abstract

A method and system for generating art uses artificial intelligence to analyze existing art forms and then creates art that deviates from the learned styles. Known art created by humans is presented in digitized form along with a style designator to a computer for analysis, including recognition of artistic elements and association of particular styles. A graphics processor generates a draft graphic image for similar analysis by the computer. The computer ranks such draft image for correlation with artistic elements and known styles. The graphics processor modifies the draft image using an iterative process until the resulting image is recognizable as art but is distinctive in style.

First claim

I claim:

1. A method for generating art comprising the steps of:
a) providing to a computer a set of digitized data characterizing a plurality of existing works of art created by human beings, the digitized data including the designation of a style associated with each such existing work of art, the computer including a memory;
b) using the computer to analyze the digitized data and corresponding style designations, including recognition of artistic elements in the existing works of art, and creating art recognition data associating particular artistic elements of existing works of art with particular style designations;
c) storing the art recognition data in the memory of the computer
d) providing a graphics processing unit to generate a graphic image;
e) providing the generated graphic image generated by the graphics processing unit to the computer;
f) using the computer to compare the generated graphic image to the stored art recognition data;
g) based upon such comparison, providing an art index which ranks the generated graphic image along a scale that ranges from being recognizable as art and not being recognizable as art;
h) based upon such comparison, providing a novelty index which ranks the generated graphic art along a scale that ranges from being within a known style for existing works of art and not being within a known style for existing works of art;
i) modifying the generated graphic image based upon the art index and the novelty index;
j) repeating steps f) through i) until the generated graphic image produces an art index indicating that the generated graphic image is recognizable as art, while also producing a novelty index indicating the generated graphic image is not within a known style for existing works of art.

2. The method recited by claim 1 wherein such method is an iterative enhancement process.

3. The method recited by claim 1 wherein the generated graphic image reached after performing such iterative enhancement process is rendered in a tangible medium.

4. The method recited by claim 1 wherein the set of digitized data comprises a plurality of paintings fixed on tangible medium.

5. The method recited by claim 1 wherein the set of digitized data comprises a plurality of sculptures fixed on tangible medium.

6. The method recited by claim 1 wherein the set of digitized data comprises a plurality of image sequences.

7. The method recited by claim 1 wherein the set of digitized data comprises a plurality of graphical designs.

8. The method recited by claim 1 wherein the set of digitized data comprises a plurality of fashion designs.

9. The method recited by claim 1 wherein the set of digitized data comprises a plurality of consumer product designs.

10. The method recited by claim 1 wherein the step of using the computer to analyze the digitized data includes the use of a neural network.

11. A computerized system for generating art comprising in combination:
a) a computer for receiving a set of digitized data characterizing a plurality of existing works of art created by human beings, the digitized data including the designation of a style associated with each such existing work of art;
b) an electronic memory coupled to the computer;
c) the computer analyzing the digitized data and corresponding style designations, including recognition of artistic elements in the existing works of art, and creating art recognition data associating particular artistic elements of existing works of art with particular style designations for storage in the electronic memory;
d) a graphics processing unit coupled to the computer for generating a graphic image, and presenting such graphic image to the computer;
f) the computer comparing the generated graphic image to the stored art recognition data, and the computer providing an art index which ranks the generated graphic image along a scale that ranges from being recognizable as art and not being recognizable as art, the computer also providing a novelty index which ranks the generated graphic art along a scale that ranges from being within a known style for existing works of art and not being within a known style for existing works of art;
g) the graphics processing unit receiving information associated with the art index and the novelty index, and being responsive thereto to generate a modified graphic image for presentation to the computer;
whereby the graphics processing unit repeatedly presents modified graphic images to the computer until a final modified graphic image presented thereby to the computer produces an art index indicating that the generated graphic image is recognizable as art, while also producing a novelty index indicating the generated graphic image is not within a known style for existing works of art.

12. The computerized system recited by claim 11 further including a printer for printing the final modified graphic image in a tangible medium.

13. The computerized system recited by claim 11 wherein the set of digitized data comprises a plurality of paintings fixed on tangible medium.

14. The computerized system recited by claim 11 wherein the set of digitized data comprises a plurality of sculptures fixed on tangible medium.

15. The computerized system recited by claim 11 wherein the set of digitized data comprises a plurality of image sequences.

16. The computerized system recited by claim 11 wherein the set of digitized data comprises a plurality of graphical designs.

17. The computerized system recited by claim 11 wherein the set of digitized data comprises a plurality of fashion designs.

18. The computerized system recited by claim 11 wherein the set of digitized data comprises a plurality of consumer product designs.

19. The computerized system recited by claim 11 further including a neural network for recognizing artistic features in works of art.

Assignees

• Elgammal Ahmed
• Univ Rutgers

Inventors

• Elgammal Ahmed

Classifications

• G10H1/0025 Primary

  Automatic or semi-automatic music composition, e.g. producing random music, applying rules from music theory or modifying a musical piece (automatically producing a series of tones G10H1/26) · CPC title

• G06V10/7753

  Incorporation of unlabelled data, e.g. multiple instance learning [MIL] · CPC title

• G06V20/00

  Scenes; Scene-specific elements (control of digital cameras H04N23/60) · CPC title

• G06V10/82

  using neural networks · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title