METHOD AND APPARATUS FOR ESTIMATING BODY SHAPE

• US 20100111370A1
• Filed: 08/14/2009
• Published: 05/06/2010
• Est. Priority Date: 08/15/2008
• Status: Active Grant

• Current Assignee: Brown University
• Original Assignee: Brown University
• Inventors: Black, Michael J., Weiss, Alexander W., St. Clair, Timothy S., Balan, Alexandru O., Loper, Matthew M., Sigal, Leonid
• Granted Patent: US 9,189,886 B2
• Time in Patent Office: Days
• US Class Current: 382/111
• CPC Class Codes: G06F 18/2321 using statistics or function optimisation, e.g. modelling of probability density functions
• CPC Class Codes: G06Q 30/0601 Electronic shopping [e-shopping]
• CPC Class Codes: G06T 17/00 Three dimensional [3D] modelling, e.g. data description of 3D objects
• CPC Class Codes: G06T 2207/10028 Range image; Depth image; 3D point clouds
• CPC Class Codes: G06T 2207/30196 Human being; Person
• CPC Class Codes: G06T 7/75 involving models
• CPC Class Codes: G06T 7/77 using statistical methods
• CPC Class Codes: G06V 10/763 Non-hierarchical techniques, e.g. based on statistics of modelling distributions
• CPC Class Codes: G06V 40/103 Static body considered as a whole, e.g. static pedestrian or occupant recognition

Abstract

A system and method of estimating the body shape of an individual from input data such as images or range maps. The body may appear in one or more poses captured at different times and a consistent body shape is computed for all poses. The body may appear in minimal tight-fitting clothing or in normal clothing wherein the described method produces an estimate of the body shape under the clothing. Clothed or bare regions of the body are detected via image classification and the fitting method is adapted to treat each region differently. Body shapes are represented parametrically and are matched to other bodies based on shape similarity and other features. Standard measurements are extracted using parametric or non-parametric functions of body shape. The system components support many applications in body scanning, advertising, social networking, collaborative filtering and Internet clothing shopping.

Claims

1. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:

obtaining input data including data representing said body in a plurality of poses; and

fitting a parametric body model of said body to the data representation of said body contained in the input data in said plurality of poses to generate multiple sets of pose parameters and at least one set of shape parameters, said at least one set of shape parameters being consistent with said plurality of poses.

2. The method of claim 1 wherein the fitting of said parametric body model to the data representation of said body includes optimizing an objective function defined at least in part by said multiple sets of pose parameters and said at least one set of shape parameters.

3. The method of claim 1 wherein said individual has one or more of an associated gender and an associated ethnicity, and wherein the fitting of said parametric body model to the data representation of said body includes optimizing an objective function defined at least in part by said multiple sets of pose parameters, said at least one set of shape parameters, and at least one specified parameter corresponding to one or more of the gender and the ethnicity of said individual.

4. The method of claim 1 wherein the obtaining of said input data includes obtaining said input data from one or more input devices selected from the group consisting of a camera, a range sensor, and an infrared sensor.

5. The method of claim 1 wherein the obtaining of said input data includes obtaining data representing at least a partially clothed body of an individual in a plurality of poses.

6. The method of claim 5 wherein the fitting of said parametric body model includes fitting a parametric body model of said at least a partially clothed body to the data representation of said at least a partially clothed body to generate parameters indicative of an estimation of a naked body shape of said at least a partially clothed body.

7. The method of claim 6 wherein the fitting of said parametric body model of said at least a partially clothed body utilizes an objective function that permits the estimated naked body shape to be substantially within the data representation of the clothed portion of the body.

8. The method of claim 1 wherein said parametric body model is a statistical parametric body model.

9. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:

obtaining input data including data representing a body of an individual in at least one pose, said body being at least partially clothed;

detecting, using the data representation of said body, at least a clothed portion of the data representation of said body, the clothed portion having an associated looseness characteristic; and

obtaining an estimation of body shape under at least the clothed portion of said body by optimizing an objective function based at least in part on the looseness characteristic.

10. The method of claim 9 wherein the optimizing of said objective function includes optimizing said objective function based at least in part on at least one set of pose parameters and a set of shape parameters, said at least one set of pose parameters and said set of shape parameters at least partially defining a parametric body model of said body.

11. The method of claim 9 wherein said individual has one or more of an associated gender and an associated ethnicity, and wherein the optimizing of said objective function includes optimizing said objective function based at least in part on one or more of the gender and the ethnicity of said individual.

12. The method of claim 9 wherein the obtaining of said input data includes obtaining said input data from one or more input devices selected from the group consisting of a camera, a range sensor, and an infrared sensor.

13. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:

obtaining one or more input images from at least one camera, said one or more input images containing depictions of said body in at least one pose, said at least one camera having an associated set of camera calibration parameters; and

obtaining an estimation of the shape of the body of said individual by optimizing an objective function based at least in part on the set of camera calibration parameters.

14. The method of claim 13 further including segmenting each of said one or more input images into a foreground region and a background region, said foreground region containing the respective depiction of said body in a respective one of said plurality of poses.

15. The method of claim 14:

wherein said objective function includes a first objective function corresponding to said foreground region of the respective input images, and a second objective function corresponding to said background region of the respective input images; and

wherein the optimizing of said objective function includes optimizing a sum of said first objective function and said second objective function.

16. The method of claim 15 wherein the optimizing of said objective function includes optimizing a weighted sum of said first objective function and said second objective function.

17. The method of claim 15:

wherein said background region corresponds to a rigid background scene;

wherein said method further includes identifying at least one feature point in the rigid background scene of each of said one or more input images;

wherein said first objective function is defined at least in part by said multiple sets of pose parameters, said set of shape parameters, and the set of camera calibration parameters; and

wherein said second objective function is defined at least in part by the set of camera calibration parameters and said at least one feature point in the rigid background scene.

18. The method of claim 14 wherein at least a portion of said background region is moving.

19. The method of claim 15:

wherein said objective function includes said first objective function, said second objective function, and at least one third objective function, the third objective function being defined at least in part by one or more of the set of camera calibration parameters; and

wherein the optimizing of said objective function includes optimizing a sum of said first objective function, said second objective function, and the third objective function.

20. The method of claim 17:

wherein said objective function includes said first objective function, said second objective function, the third objective function, and at least one fourth objective function, the fourth objective function being defined at least in part by an amount of variation in at least one of the set of camera calibration parameters; and

wherein the optimizing of said objective function includes optimizing a sum of said first objective function, said second objective function, the third objective function, and the fourth objective function.

21. The method of claim 18 wherein the optimizing of said objective function includes optimizing a weighted sum of said first objective function, said second objective function, the third objective function, and the fourth objective function.

22. The method of claim 17:

wherein said objective function includes said first objective function, said second objective function, the third objective function, the fourth objective function, and at least one fifth objective function, the fifth objective function being defined at least in part by an amount of variation in at least one of said plurality of poses; and

wherein the optimizing of said objective function includes optimizing a sum of said first objective function, said second objective function, the third objective function, the fourth objective function, and the fifth objective function.

23. The method of claim 22 wherein the optimizing of said objective function includes optimizing a weighted sum of said first objective function, said second objective function, the third objective function, the fourth objective function, and the fifth objective function.

24. The method of claim 13:

wherein said objective function includes said first objective function, and at least one second objective function, said at least one second objective function being defined at least in part by an amount of variation in at least one of said plurality of poses; and

wherein the optimizing of said objective function includes optimizing a sum of said first objective function and said at least one second objective function.

25. The method of claim 14:

wherein each of said one or more input images depicts said body disposed in front of a first surface within said background region, said first surface having a predetermined multi-color pattern disposed thereon; and

wherein said method further includes calculating at least one of the set of camera calibration parameters using said first surface.

26. The method of claim 14:

wherein each of said one or more input images depicts said body disposed in front of a first surface and a second surface, each of the first and second surfaces being within said background region, each of the first and second surfaces having a predetermined multi-color pattern disposed thereon; and

wherein said method further includes calculating at least one of the set of camera calibration parameters based on at least one homography between the camera and at least one of the first and second surfaces.

27. The method of claim 26:

wherein each of the first and second surfaces having said predetermined multi-color pattern disposed thereon has an associated color distribution; and

wherein the segmenting of each of said one or more input images into said foreground region and said background region includes segmenting each of said one or more input images based at least in part on the color distribution associated with at least one of the first and second surfaces.

28. The method of claim 13 wherein said at least one camera is moving and said body is still.

29. The method of claim 13 wherein said at least one camera is still and said body is moving.

30. The method of claim 13 wherein both of said at least one camera and said body are moving.

31. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:

obtaining input data including data representing said body in at least one pose, said body having at least one associated attribute, the at least one attribute of said body being subject to at least one constraint;

fitting a parametric body model of said body to the data representation of said body while substantially satisfying said at least one constraint on the at least one attribute of said body, said parametric body model being defined by a plurality of parameters including at least one set of pose parameters and a set of shape parameters,

whereby said parametric body model has an associated shape that is representative of the shape of the body of said individual.

32. The method of claim 31 wherein said at least one constraint is operative to constrain the at least one attribute of said body to at least one fixed value.

33. The method of claim 32:

wherein there is a predetermined linear relationship between the at least one attribute of said body and said set of shape parameters; and

wherein said method further includes calculating the at least one attribute of said body from said set of shape parameters based on said predetermined linear relationship.

34. The method of claim 31 wherein the fitting of said parametric body model to the data representation of said body includes optimizing a first objective function defined at least in part by said at least one set of pose parameters and said set of shape parameters.

35. The method of claim 34 wherein the fitting of said parametric body model to the data representation of said body includes optimizing a sum of said first objective function and at least one second objective function, the second objective function being defined at least in part by the at least one attribute of said body.

36. The method of claim 34 wherein the fitting of said parametric body model to the data representation of said body includes optimizing a sum of said first objective function and at least one second objective function, the second objective function being defined at least in part by the at least one attribute of said body and a third function, said third function being defined by said set of shape parameters.

37. The method of claim 34 wherein the fitting of said parametric body model to the data representation of said body includes optimizing a sum of said first objective function and at least one second objective function, the second objective function being defined at least in part by the at least one attribute of said body and a third function, said third function being defined by said set of shape parameters, said third function being operative to predict at least one of the at least one attribute of said body based on said set of shape parameters.

38. A computer implemented method of obtaining camera calibration and image segmentation, comprising the steps of:

obtaining, from at least one camera, image data associated with at least one image of an object in front of at least one surface, said at least one surface having a predetermined multi-chroma pattern with predetermined pattern dimensions;

in a first processing step, processing the image data from the at least one image of said object to obtain one or more intrinsic or extrinsic parameters associated with the camera; and

in a second processing step, utilizing the same image data associated with the at least one image of said object, performing segmentation to segment a foreground image of the object from said at least one surface.

39. The method of claim 38 further including generating an estimation of one or more of object shape and pose utilizing the intrinsic or extrinsic parameters and the segmented foreground image obtained in the first and second processing steps, respectively.

40. The method of claim 38 wherein the image has an associated foreground region and an associated background region, wherein the image includes a plurality of pixels, and wherein said method further includes:

calculating, using said predetermined multi-chroma pattern disposed on said at least one surface, a color distribution for said at least one surface;

determining, based on the respective color distribution for said at least one surface, a likelihood that each of said plurality of pixels in the image depicts a portion of said at least one surface;

for each pixel having a low likelihood of depicting a portion of said at least one surface, segmenting the respective pixel to said foreground region of the image; and

otherwise, segmenting the respective pixel to said background region of the image.

41. The method of claim 40 further including calculating, from the color distribution for said at least one surface, a probability image for said at least one surface, wherein each probability image is operative to provide an indication of a probability of one of said plurality of pixels being generated from the color distribution of said at least one surface.

42. The method of claim 41 further including calculating a per-pixel maximum of the probability images.

43. The method of claim 42 wherein the segmenting of the respective pixels includes thresholding said per-pixel maximum of the probability images.

44. The method of claim 42 wherein said predetermined multi-chroma pattern is a checkerboard pattern.

45. The method of claim 44 wherein said checkerboard pattern is blue and green.

46. The method of claim 38 wherein said object at least partially occludes said at least one surface.

47. The method of claim 46 wherein said object is selected from the group consisting of an inanimate object, a person, and an animal.

48. A computer implemented method of estimating a measurement associated with a body of an individual, said body having an associated shape and at least one associated attribute, comprising the steps of:

obtaining a mapping of a representation of the shape of the body of said individual to the at least one attribute of said body;

obtaining body shape information pertaining to the shape of the body of said individual; and

estimating at least one characteristic associated with the body of said individual using the body shape information for said individual and the mapping of said representation of the shape of the body of said individual to the at least one attribute of said body.

49. The method of claim 48:

wherein the body shape information for said individual includes a set of shape parameters defining, at least in part, a parametric body model for the body of said individual; and

wherein the estimating of said at least one characteristic includes estimating said at least one characteristic using said set of shape parameters for said parametric body model.

50. The method of claim 49 wherein said parametric body model has a plurality of vertex coordinates associated therewith, and wherein the obtaining of the mapping of the representation of the shape of said body to the at least one attribute of said body includes identifying at least a portion of said plurality of vertex coordinates for said parametric body model, and learning a function for performing a parametric or non-parametric mapping of the vertex coordinates to the at least one attribute of said body.

51. The method of claim 50 wherein the learning of said function for mapping the vertex coordinates to the at least one attribute includes learning said function using a linear regression technique.

52. The method of claim 49 wherein said parametric body model has a plurality of measurements associated therewith, and wherein the obtaining of the mapping of the representation of the shape of said body to the at least one attribute of said body includes identifying at least a portion of said plurality of measurements for said parametric body model, and learning a function for performing a parametric or non-parametric mapping of the measurements to point locations on said parametric body model.

53. The method of claim 52 wherein the learning of said function for mapping the measurements to the point locations includes learning said function using a linear regression technique.

54. The method of claim 48 wherein the obtaining of the mapping of the representation of the shape of the body of said individual to the at least one attribute of said body includes a set of shape parameters

55. The method of claim 48 wherein the estimating of said at least one characteristic is performed in association with one or more of a virtual try-on of clothing by said individual, a purchase of clothing for said individual, and an animation of a representation of said individual.

56. A method of matching attributes associated with bodies of a plurality of individuals, comprising the steps of:

obtaining input data representing a body of a first individual;

generating, using said input data, information defining a parametric body model of said first individual, the parametric body model of said first individual having at least one associated attribute;

accessing, from a database, information defining at least one parametric body model of at least one second individual, the parametric body model of the second individual having at least one associated attribute;

computing a measure of similarity between the at least one attribute of the parametric body model of said first individual and the at least one attribute of said at least one parametric body model of said at least one second individual; and

generating a value indicative of the similarly between the attributes of the parametric body models of the first and second individuals based at least in part on the computed measure of similarity.

57. The method of claim 56 wherein the at least one attribute corresponds to at least one shape coefficient.

58. The method of claim 56:

wherein each of the parametric body model of said first individual and the at least one parametric body model of said at least one second individual has a plurality of associated vertex locations;

wherein each of the parametric body model of said first individual and the at least one parametric body model of said at least one second individual defines a mesh;

wherein each of the at least one parametric body model of said at least one second individual is in alignment with the mesh of the parametric body model of said first individual; and

wherein the computing of said measure of similarity includes computing distances between selected ones of the vertex locations of the parametric body model of said first individual and corresponding ones of the vertex locations of the at least one parametric body model of said at least one second individual.

59. The method of claim 56:

wherein the accessing of said information includes accessing information defining a plurality of parametric body models of a plurality of second individuals, respectively;

wherein the computing of said measure of similarity includes computing multiple measures of similarity between the at least one attribute of the parametric body model of said first individual and the attributes of the respective parametric body models of the plurality of second individuals; and

wherein said method further includes generating, using said multiple measures of similarity, a ranking of the similarity between the at least one attribute of the parametric body model of said first individual and the attributes of the respective parametric body models of the plurality of second individuals.

60. The method of claim 56 further including estimating at least one measurement associated with the body of said first individual based at least in part on the computed measure of similarity between the at least one attribute of the parametric body model of said first individual and the at least one attribute of the parametric body model of the second individual.

61. The method of claim 60 wherein the estimating of said at least one measurement associated with the body of said first individual is performed in association with a purchase of clothing for said first individual.

62. The method of claim 56 further including generating a recommendation pertaining to one or more of a clothing attribute, a body measurement attribute, and a body shape attribute based on said value indicative of the similarly between the respective attributes.

63. The method of claim 56 further including executing at least one application that uses said value in a decisional process.

64. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:

obtaining a database containing body shape information, body pose information, and feature information, said body shape information pertaining to shapes of a plurality of bodies of individuals, said body pose information pertaining to poses associated with said plurality of bodies, said feature information comprising at least one transformation of real or simulated sensor data for each of said plurality of bodies;

generating, using said body shape information, said body pose information, and said feature information contained in said database, a mapping of feature information to body shape and pose information;

obtaining input sensor data representing the body of said individual;

obtaining feature information for said individual from said input sensor data, said feature information for said individual comprising at least one transformation of said input sensor data for the body of said individual; and

estimating, using the feature information for said individual and said mapping of feature information to body shape and pose information, the shape of the body of said individual.

65. The method of claim 64 further including:

generating information defining a 3-dimensional parametric body model having an associated shape corresponding to the shape of the body of said individual;

producing a 2-dimensional silhouette of said parametric body model of the body of said individual, said silhouette having an associated contour; and

wherein the obtaining of the feature information for said individual includes obtaining the feature information for said individual from the contour of the 2-dimensional silhouette.

66. The method of claim 64 further including:

obtaining input data representing the body of said individual in 3-dimensions; and

wherein the obtaining of the feature information for said individual includes obtaining the feature information for said individual from said data representing the body of said individual in 3-dimensions.

67. The method of claim 66 further including generating, using said input data, information defining a 3-dimensional parametric body model having an associated shape corresponding to the shape of the body of said individual.

68. A computer implemented method of determining an indication of a product associated with a human body, comprising the steps of:

obtaining a representation of at least one aspect of a body of an individual;

comparing the representation of the at least one aspect of the body of said individual to at least one representation of at least one aspect of each of a plurality of bodies of different individuals contained in a database, wherein each representation associated with each of the different individuals has a product associated therewith; and

obtaining a ranking of similarity of the representation of the at least one aspect of the body of said individual to the representation of the at least one aspect of the bodies of said plurality of different individuals.

69. The method of claim 68 further including selecting, based upon the ranking of similarity, at least one advertisement for at least one product.

70. The method of claim 68 further including displaying, based upon the ranking of similarity, a graphical image of at least one product associated with one of the different individuals.

71. The method of claim 68:

wherein each of the body of said individual and the bodies of the different individuals is represented by a parametric body model;

wherein the comparing of said representation of the at least one aspect of the body of said individual to said at least one representation of at least one aspect of each of said plurality of bodies of different individuals includes computing multiple measures of similarity between the shape of the parametric body model of said individual and the shapes of the parametric body models of the respective different individuals; and

wherein said method further includes generating, using said multiple measures of similarity, a ranking of the similarity between the shape of the parametric body model of said individual and the shapes of the parametric body models of the respective different individuals.

72. The method of claim 71 further including outputting an indication of said ranking on a display.

73. The method of claim 71 further including estimating at least one measurement associated with the body of said individual based at least in part on the computed measures of similarity between the shape of the parametric body model of said individual and the shapes of the parametric body models of the respective different individuals.

74. The method of claim 73 further including outputting an indication of said at least one measurement on a display or using said at least one measurement as an input to at least one application.

75. A system for estimating a shape of a body of an individual, comprising:

an input device operative to obtain input data including data representing said body in a plurality of poses; and

at least one processor operative to execute at least one program out of at least one memory to fit a parametric body model of said body to the data representation of said body contained in the input data in said plurality of poses to generate multiple sets of pose parameters and at least one set of shape parameters, said at least one set of shape parameters being consistent with said plurality of poses.

76. A system for estimating a shape of a body of an individual, comprising:

an input device operative to obtain input data including data representing a body of an individual in at least one pose, said body being at least partially clothed; and

at least one processor operative to execute at least one program out of at least one memory to detect, using the data representation of said body, at least a clothed portion of the data representation of said body, the clothed portion having an associated looseness characteristic, and to generate an estimation of body shape under at least the clothed portion of said body by optimizing an objective function based at least in part on the looseness characteristic.

77. A system for estimating a shape of a body of an individual, comprising:

an input device operative to obtain a plurality of input images from at least one camera, said plurality of input images containing depictions of said body in at least one pose, said at least one camera having an associated set of camera calibration parameters; and

at least one processor operative to execute at least one program out of at least one memory to generate an estimation of the shape of the body of said individual by optimizing an objective function based at least in part on the set of camera calibration parameters.

78. A system for estimating a shape of a body of an individual, comprising:

an input device operative to obtain input data including data representing said body in at least one pose, said body having at least one associated attribute, the at least one attribute of said body being subject to at least one constraint; and

at least one processor operative to execute at least one program out of at least one memory to fit a parametric body model of said body to the data representation of said body while substantially satisfying said at least one constraint on the at least one attribute of said body, said parametric body model being defined by a plurality of parameters including at least one set of pose parameters and a set of shape parameters,

whereby said parametric body model has an associated shape that is representative of the shape of the body of said individual.

79. A system for obtaining camera calibration and image segmentation, comprising:

at least one surface having a predetermined multi-chroma pattern with predetermined pattern dimensions;

an input device operative to obtain, from at least one camera, image data associated with at least one image of an object in front of said at least one surface; and

at least one processor operative to execute at least one program out of at least one memory;

to process the image data from the at least one image of said object to obtain one or more intrinsic or extrinsic parameters associated with the camera; and

utilizing the same image data associated with the at least one image of said object, to perform segmentation to segment a foreground image of the object from said at least one surface.

80. The system of claim 79 wherein said at least one processor is further operative to execute said at least one program out of said at least one memory to generate an estimation of object shape utilizing the intrinsic or extrinsic parameters and the segmented foreground image obtained in the first and second processing steps, respectively.

81. A system for estimating a measurement associated with a body of an individual, said body having an associated shape and at least one associated attribute, comprising:

at least one processor operative to execute at least one program out of at least one memory;

to receive body shape information pertaining to the shape of the body of said individual; and

to estimate at least one characteristic associated with the body of said individual using the body shape information for said individual and a mapping of a representation of the shape of the body of said individual to the at least one attribute of said body.

82. A system for matching attributes associated with bodies of a plurality of individuals, comprising:

at least one processor operative to execute at least one program out of at least one memory;

to receive input data representing a body of a first individual;

to generate, using said input data, information defining a parametric body model of said first individual, the parametric body model of said first individual having at least one associated attribute;

to access, from a database, information defining at least one parametric body model of at least one second individual, the parametric body model of the second individual having at least one associated attribute;

to compute a measure of similarity between the at least one attribute of the parametric body model of said first individual and the at least one attribute of said at least one parametric body model of said at least one second individual; and

to generate a value indicative of the similarly between the attributes of the parametric body models of the first and second individuals based at least in part on the computed measure of similarity.

83. A system for estimating a shape of a body of an individual, comprising:

a database containing body shape information, body pose information, and feature information, said body shape information pertaining to shapes of a plurality of bodies of individuals, said body pose information pertaining to poses associated with said plurality of bodies, said feature information comprising at least one transformation of sensor data for each of said plurality of bodies;

at least one input device operative to obtain input sensor data representing the body of said individual; and

at least one processor operative to execute at least one program out of at least one memory;

to generate feature information for said individual from said input sensor data, said feature information for said individual comprising at least one transformation of said input sensor data for the body of said individual;

to generate, using said body shape information, said body pose information, and said feature information contained in said database, a mapping of feature information to body shape and pose information; and

to estimate, using the feature information for said individual and said mapping of feature information to body shape and pose information, the shape of the body of said individual.

84. A system for determining an indication of a product associated with a human body, comprising:

at least one processor operative to execute at least one program out of at least one memory;

to receive a representation of at least one aspect of a body of an individual;

to compare the representation of the at least one aspect of the body of said individual to at least one representation of at least one aspect of each of a plurality of bodies of different individuals contained in a database, wherein each representation associated with each of the different individuals has a product associated therewith; and

to generate a ranking of similarity of the representation of the at least one aspect of the body of said individual to the representation of the at least one aspect of the bodies of said plurality of different individuals.

85. A computer implemented method of determining attribute similarities, comprising the steps of:

receiving input data representing a body of a first individual;

generating, using said input data, information defining a parametric body model of said first individual, the parametric body model of said first individual having at least one associated attribute;

accessing, from a database, information corresponding to at least one attribute specification; and

in a computing step, computing a score indicative of a similarity between the at least one attribute associated with the first individual and each of the at least one attribute specification.

86. A system for determining attribute similarities, comprising:

At least one processor operative to execute at least one program out of at least one memory;

to receive input data representing a body of a first individual;

to generate, using said input data, information defining a parametric body model of said first individual, the parametric body model of said first individual having at least one associated attribute;

to access, from a database, information corresponding to at least one attribute specification; and

to compute a score indicative of a similarity between the at least one attribute associated with the first individual and each of the at least one attribute specification.