Granted Patent B2

US 12,633,076 · App. 19/098,005 · Granted May 19, 2026

Inventors: Caleb Bridges King (Roanoke, VA); Shiraz Sky Alibhai (Durham, NC)

Assignee: JMP STATISTICAL DISCOVERY LLC

G06T19/20G06T7/543G06T7/74G06V10/764G06V10/7715G06F3/04845G06T2200/24G06T2207/20081G06T2207/20084G06T2219/2016G06T2219/2021

Abstract

Claims (116)

1 . A computer-program product comprising a non-transitory machine-readable storage medium having instructions embodied therewith, the instructions operable to cause a computing system to:

display, in a graphical user interface, a graphical representation of position information,

wherein the position information defines a first representation of a shape of at least two dimensions, and

wherein the first representation comprises a first coordinate set describing at least one feature of the shape in the at least two dimensions;

obtain a computer model that predicts an initial response according to the first representation input to the computer model, wherein the initial response indicates a likelihood of classifying the shape according to multiple response outcomes;

display, in the graphical user interface, in proximity to the first coordinate set of the first representation, a prediction indicator predicting an extent of change to the likelihood of classifying the shape given changes in the first coordinate set;

receive, via the graphical user interface, a manipulation of the graphical representation that changes the position information to define a second representation of the shape of the at least two dimensions; and

generate an updated response according to the second representation input to the computer model.

2 . The computer-program product of claim 1 ,

wherein the initial response and the updated response for the computer model both indicate responses of the computer model for the shape of the at least two dimensions; and

wherein the initial response and the updated response are different responses.

3 . The computer-program product of claim 1 , wherein the prediction indicator is a sensitivity indicator that predicts, based on a size, direction, or both of the sensitivity indicator displayed in the graphical user interface, the extent of change to the likelihood of classifying the shape.

4 . The computer-program product of claim 1 ,

wherein the graphical user interface displays the multiple response outcomes; and

wherein the instructions are operable to further cause the computing system to display, in the graphical user interface, in proximity to the first coordinate set of the first representation, multiple prediction indicators with at least one predictor indicator predicting a respective extent of change to the likelihood of classifying the shape for each of the multiple response outcomes, wherein the multiple prediction indicators comprise the prediction indicator.

5 . The computer-program product of claim 4 , wherein the instructions are operable to further cause the computing system to:

receive a selection of a subset of the multiple response outcomes; and

display, in proximity to the first coordinate set of the first representation, only prediction indicators of the multiple prediction indicators that correspond to the selection of the subset of the multiple response outcomes.

6 . The computer-program product of claim 1 ,

wherein the initial response is a categorical response type with at least two categories for the multiple response outcomes; and

wherein the instructions are operable to further cause the computing system to display, in the graphical user interface, in proximity to the first coordinate set of the first representation, multiple prediction indicators with at least one predictor indicator for each of the at least two categories, wherein the multiple prediction indicators comprise the prediction indicator.

7 . The computer-program product of claim 1 ,

wherein the first coordinate set comprises a value for each dimension of the shape; and

wherein the instructions are operable to further cause the computing system to display, in the graphical user interface, in proximity to the first coordinate set of the first representation, multiple prediction indicators with at least one predictor indicator for changes to each value of the first coordinate set, wherein the multiple prediction indicators comprise the prediction indicator.

8 . The computer-program product of claim 1 ,

wherein the graphical representation depicts unique positions within the graphical representation based on the position information; and

wherein the instructions are operable to further cause the computing system to receive the manipulation by receiving an indication to move one or more of the unique positions within the graphical representation.

9 . The computer-program product of claim 1 ,

wherein the graphical representation depicts unique positions within the graphical representation based on the position information; and

wherein the instructions are operable to further cause the computing system to receive the manipulation by receiving an indication to move a set of the unique positions within the graphical representation.

10 . The computer-program product of claim 1 ,

wherein the first representation comprises a second coordinate set describing another feature of the shape; and

wherein the instructions are operable to further cause the computing system to:

receive the manipulation of the graphical representation by receiving an indication to change one or more coordinates in the first coordinate set to a user-defined coordinate set;

responsive to the manipulation of the graphical representation:

generate one or more coordinates updating the second coordinate set to a computer-generated coordinate set; and

generate the updated response according to the computer model, the user-defined coordinate set, and the computer-generated coordinate set.

11 . The computer-program product of claim 10 , wherein the instructions are operable to further cause the computing system to receive the manipulation of the graphical representation by:

receiving an indication to change the initial response to the updated response; and

generating a computer-generated manipulation of the graphical representation to achieve the updated response.

12 . The computer-program product of claim 1 ,

wherein the initial response is a categorical response type; and

wherein the initial response indicates a classification of the shape by assessing a likelihood of the shape being a particular one of multiple object categories; and

wherein the instructions are operable to further cause the computing system to generate the updated response by changing the classification of the shape into a different one of the multiple object categories.

13 . The computer-program product of claim 1 ,

wherein the initial response is a continuous response type; and

wherein the initial response indicates the shape of the at least two dimensions by:

quantifying a property of an object along a range based on the shape; and

generating a likelihood of an outcome for a continuous response type.

14 . The computer-program product of claim 1 ,

wherein the instructions are operable to further cause the computing system to obtain the computer model by training the computer model on multiple objects preclassified into object categories; and

wherein the shape of the at least two dimensions is for an object that is not a member of the multiple objects.

15 . The computer-program product of claim 1 , wherein the instructions are operable to further cause the computing system to:

receive an indication to display the initial response or the updated response in proximity to the graphical representation; and

display, based on the indication, the initial response or the updated response in proximity to the graphical representation.

16 . The computer-program product of claim 1 , wherein the instructions are operable to further cause the computing system to display the first representation in the graphical user interface by:

obtaining a detected object;

applying an outline to the detected object;

generating coordinate sets for several points along the outline, or synthesized from the several points along the outline, wherein each coordinate set defines a position for a respective point along the outline within the at least two dimensions; and

displaying one or more of the coordinate sets in the graphical representation.

17 . The computer-program product of claim 1 , wherein the instructions are operable to further cause the computing system to display the first representation in the graphical user interface by:

obtaining a detected object;

determining representative coordinate sets for the detected object according to an object template defining landmark features for detected objects; and

displaying one or more of the representative coordinate sets in the graphical representation.

18 . The computer-program product of claim 1 ,

wherein the shape of the at least two dimensions is generated from an image of a detected object in a real-world environment;

wherein the instructions are operable to further cause the computing system to display, in the graphical user interface, the first representation of the shape of the at least two dimensions by aligning the shape of the at least two dimensions to a reference frame for the graphical representation; and

wherein aligning the shape to the reference frame for the graphical representation comprises one or more of:

reducing translation variation in the shape;

rotating the shape; and

scaling the shape.

19 . The computer-program product of claim 1 ,

wherein the shape of the at least two dimensions is generated from a simulated object in a simulated environment; and

wherein the instructions are operable to further cause the computing system to receive the manipulation of the graphical representation according to a change for the simulated object or the simulated environment.

20 . The computer-program product of claim 1 ,

wherein the computer model predicts the initial response according to the first representation and at least one other factor other than the shape of the at least two dimensions; and

wherein instructions are operable to further cause the computing system to generate the updated response according to the computer model, the manipulation of the graphical representation and an indication to change one or more of the at least one other factor.

21 . The computer-program product of claim 1 ,

wherein the shape of the at least two dimensions has at least three dimensions;

wherein the position information defines the first representation by comprising the first coordinate set describing the at least one feature of the shape in the at least three dimensions;

wherein the graphical representation represents the first coordinate set; and

wherein the instructions are operable to further cause the computing system to:

receive constraints related to limits for changes to the position information; and

display, in the graphical user interface, in proximity to the first coordinate set of the graphical representation, the prediction indicator predicting the extent of change to the likelihood of classifying the shape given the changes in the first coordinate set accounting for the constraints.

22 . The computer-program product of claim 1 ,

wherein the graphical user interface is an interactive graphical user interface comprising a shape profiler;

wherein the shape profiler comprises the graphical representation in proximity to a response viewer; and

wherein the instructions are operable to further cause the computing system to generate the updated response by causing the response viewer to change a display of the initial response to the updated response in the interactive graphical user interface.

23 . A computer-implemented method comprising:

displaying, in a graphical user interface, a graphical representation of position information,

wherein the position information defines a first representation of a shape of at least two dimensions, and

wherein the first representation comprises a first coordinate set describing at least one feature of the shape in the at least two dimensions;

obtaining a computer model that predicts an initial response according to the first representation input to the computer model, wherein the initial response indicates a likelihood of classifying the shape according to multiple response outcomes;

displaying, in the graphical user interface, in proximity to the first coordinate set of the first representation, a prediction indicator predicting an extent of change to the likelihood of classifying the shape given changes in the first coordinate set;

receiving, via the graphical user interface, a manipulation of the graphical representation that changes the position information to define a second representation of the shape of the at least two dimensions; and

generating an updated response according to the second representation input to the computer model.

24 . The computer-implemented method of claim 23 ,

wherein the initial response is a categorical response type with at least two categories for the multiple response outcomes; and

wherein the computer-implemented method further comprises displaying, in the graphical user interface, in proximity to the first coordinate set of the first representation, multiple prediction indicators with at least one predictor indicator for each of the at least two categories, wherein the multiple prediction indicators comprise the prediction indicator.

25 . The computer-implemented method of 23 ,

wherein the first coordinate set comprises a value for each dimension of the shape; and

wherein the computer-implemented method further comprises displaying, in the graphical user interface, in proximity to the first coordinate set of the first representation, multiple prediction indicators with at least one predictor indicator for changes to each value of the first coordinate set, wherein the multiple prediction indicators comprise the prediction indicator.

26 . The computer-implemented method of 23 ,

wherein the graphical user interface displays the multiple response outcomes; and

wherein the computer-implemented method further comprises displaying in the graphical user interface, in proximity to the first coordinate set of the first representation, multiple prediction indicators with at least one predictor indicator predicting a respective extent of change to the likelihood of classifying the shape for each of the multiple response outcomes, wherein the multiple prediction indicators comprise the prediction indicator.

27 . The computer-implemented method of 23 ,

wherein the graphical representation depicts unique positions within the graphical representation based on the position information; and

wherein the computer-implemented method further comprises receiving the manipulation by receiving an indication to move one or more of the unique positions within the graphical representation.

28 . A computing device comprising processor and memory, the memory containing instructions executable by the processor wherein the computing device is configured to:

display, in a graphical user interface, a graphical representation of position information,

wherein the position information defines a first representation of a shape of at least two dimensions, and

wherein the first representation comprises a first coordinate set describing at least one feature of the shape in the at least two dimensions;

obtain a computer model that predicts an initial response according to the first representation input to the computer model, wherein the initial response indicates a likelihood of classifying the shape according to multiple response outcomes;

display, in the graphical user interface, in proximity to the first coordinate set of the first representation, a prediction indicator predicting an extent of change to the likelihood of classifying the shape given changes in the first coordinate set;

receive, via the graphical user interface, a manipulation of the graphical representation that changes the position information to define a second representation of the shape of the at least two dimensions; and

generate an updated response according to the second representation input to the computer model.

Assignments (1)

Continuity (3)

References Cited (22)