Arbitrary view generation

1. A system, comprising:

a database configured to store sets of images associated with corresponding assets; and

a processor configured to generate an image comprising a requested view of a scene using at least one or more subsets of images associated with one or more assets comprising the scene that are stored in the database, wherein the processor is configured to transform at least some images of the at least one or more subsets to a perspective corresponding to the requested view when generating the image comprising the requested view of the scene.

2. The system of claim 1 , wherein camera information is stored for a perspective of an asset stored in the database.

3. The system of claim 1 , wherein camera information is stored for a perspective of an asset stored in the database and includes one or more of: position, orientation, rotation, angle, depth, focal length, aperture, and zoom level.

4. The system of claim 1 , wherein lighting information is stored for a perspective of an asset stored in the database.

5. The system of claim 1 , wherein each of the sets of images comprises reference views and images encoding metadata.

6. The system of claim 1 , wherein xyz coordinates are stored for a pixel of a perspective of an asset stored in the database.

7. The system of claim 1 , wherein uv coordinates are stored for a pixel of a perspective of an asset stored in the database.

8. The system of claim 1 , wherein surface normal vectors are stored for a pixel of a perspective of an asset stored in the database.

9. The system of claim 1 , wherein texture values are stored for a pixel of a perspective of an asset stored in the database.

10. The system of claim 1 , wherein one or more reference views and associated metadata of assets stored in the database are generated from corresponding three-dimensional models of those assets.

11. The system of claim 1 , wherein the requested view comprises one or more of arbitrary camera characteristics, lighting, and textures.

12. The system of claim 1 , wherein at least some of the sets of images comprise training data sets for machine learning.

13. The system of claim 12 , wherein the training data sets are tagged with metadata.

14. The system of claim 12 , wherein the training data sets are associated with a prescribed modeled environment.

15. The system of claim 12 , wherein the processor is further configured to learn metadata values associated with the training data sets using a neural network.

16. The system of claim 15 , wherein the processor is further configured to predict metadata values for an image for which such metadata values are unknown using the neural network.

17. The system of claim 16 , wherein the processor is further configured to store the image and predicted metadata values of the image in the database.

18. The system of claim 15 , wherein the processor is further configured to predict metadata values for images for which such metadata values are unknown and wherein the images comprise photographs captured in a physical environment that is modeled by a modeled environment used to render the training data sets.

19. A method, comprising:

storing sets of images associated with corresponding assets in a database; and

generating an image comprising a requested view of a scene using at least one or more subsets of images associated with one or more assets comprising the scene that are stored in the database, wherein at least some images of the at least one or more subsets are transformed to a perspective corresponding to the requested view when generating the image comprising the requested view of the scene.

20. The method of claim 19 , wherein camera information is stored for a perspective of an asset stored in the database.

21. The method of claim 19 , wherein camera information is stored for a perspective of an asset stored in the database and includes one or more of: position, orientation, rotation, angle, depth, focal length, aperture, and zoom level.

22. The method of claim 19 , wherein lighting information is stored for a perspective of an asset stored in the database.

23. The method of claim 19 , wherein each of the sets of images comprises reference views and images encoding metadata.

24. The method of claim 19 , wherein xyz coordinates are stored for a pixel of a perspective of an asset stored in the database.

25. The method of claim 19 , wherein uv coordinates are stored for a pixel of a perspective of an asset stored in the database.

26. The method of claim 19 , wherein surface normal vectors are stored for a pixel of a perspective of an asset stored in the database.

27. The method of claim 19 , wherein texture values are stored for a pixel of a perspective of an asset stored in the database.

28. The method of claim 19 , wherein one or more reference views and associated metadata of assets stored in the database are generated from corresponding three-dimensional models of those assets.

29. The method of claim 19 , wherein the requested view comprises one or more of arbitrary camera characteristics, lighting, and textures.

30. The method of claim 19 , wherein at least some of the sets of images comprise training data sets for machine learning.

31. The method of claim 30 , wherein the training data sets are tagged with metadata.

32. The method of claim 30 , wherein the training data sets are associated with a prescribed modeled environment.

33. The method of claim 30 , further comprising learning metadata values associated with the training data sets using a neural network.

34. The method of claim 33 , further comprising predicting metadata values for an image for which such metadata values are unknown using the neural network.

35. The method of claim 34 , further comprising storing the image and predicted metadata values of the image in the database.

36. The method of claim 33 , further comprising predicting metadata values for images for which such metadata values are unknown and wherein the images comprise photographs captured in a physical environment that is modeled by a modeled environment used to render the training data sets.

37. A computer program product, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:

storing sets of images associated with corresponding assets in a database; and

generating an image comprising a requested view of a scene using at least one or more subsets of images associated with one or more assets comprising the scene that are stored in the database, wherein at least some images of the at least one or more subsets are transformed to a perspective corresponding to the requested view when generating the image comprising the requested view of the scene.

38. The computer program product of claim 37 , wherein camera information is stored for a perspective of an asset stored in the database.

39. The computer program product of claim 37 , wherein camera information is stored for a perspective of an asset stored in the database and includes one or more of: position, orientation, rotation, angle, depth, focal length, aperture, and zoom level.

40. The computer program product of claim 37 , wherein lighting information is stored for a perspective of an asset stored in the database.

41. The computer program product of claim 37 , wherein each of the sets of images comprises reference views and images encoding metadata.

42. The computer program product of claim 37 , wherein xyz coordinates are stored for a pixel of a perspective of an asset stored in the database.

43. The computer program product of claim 37 , wherein uv coordinates are stored for a pixel of a perspective of an asset stored in the database.

44. The computer program product of claim 37 , wherein surface normal vectors are stored for a pixel of a perspective of an asset stored in the database.

45. The computer program product of claim 37 , wherein texture values are stored for a pixel of a perspective of an asset stored in the database.

46. The computer program product of claim 37 , wherein one or more reference views and associated metadata of assets stored in the database are generated from corresponding three-dimensional models of those assets.

47. The computer program product of claim 37 , wherein the requested view comprises one or more of arbitrary camera characteristics, lighting, and textures.

48. The computer program product of claim 37 , wherein at least some of the sets of images comprise training data sets for machine learning.

49. The computer program product of claim 48 , wherein the training data sets are tagged with metadata.

50. The computer program product of claim 48 , wherein the training data sets are associated with a prescribed modeled environment.

51. The computer program product of claim 48 , further comprising computer instructions for learning metadata values associated with the training data sets using a neural network.

52. The computer program product of claim 51 , further comprising computer instructions for predicting metadata values for an image for which such metadata values are unknown using the neural network.

53. The computer program product of claim 52 , further comprising computer instructions for storing the image and predicted metadata values of the image in the database.

54. The computer program product of claim 51 , further comprising computer instructions for predicting metadata values for images for which such metadata values are unknown and wherein the images comprise photographs captured in a physical environment that is modeled by a modeled environment used to render the training data sets.