Derek J Shiell

US Patent:

8144976, Mar 27, 2012

Filed:

Dec 6, 2010

Appl. No.:

12/961350

Inventors:

Derek Shiell - Los Angeles CA, US
Jing Xiao - Cupertino CA, US

Assignee:

Seiko Epson Corporation - Tokyo

International Classification:

G06K 9/00
G06T 17/00
G06F 19/00

US Classification:

382159, 345420, 700 96

Abstract:

An Active Appearance Model AAM is trained using expanded library having examples of true outlier images. The AAM creates a first statistical fitting pair (a model image of the class of object and corresponding statistical model fitting) using characteristic features drawn only from the expanded library. All images within the expanded library that the first statistical fitting pair cannot align are collected into a smaller, second library of true outlier cases. A second statistical fitting pair is created using characteristic features drawn only from the second library, and samples not aligned by the second statistical fitting pair are collected into a still smaller, third library. This process is repeated until a desired percentage of all the images within the initial, expanded library have been aligned. In operation, the AAM applies each of its created statistical fitting pairs, in turn, until it has successfully aligned a submitted test image, or until a stop criterion has been reached.

US Patent:

8204301, Jun 19, 2012

Filed:

Feb 25, 2009

Appl. No.:

12/392820

Inventors:

Jing Xiao - Cupertino CA, US
Derek Shiell - Palo Alto CA, US

Assignee:

Seiko Epson Corporation - Tokyo

International Classification:

G06K 9/00
G06T 15/00
G06T 17/00

US Classification:

382154, 345419, 345420

Abstract:

Aspects of the present invention include systems and methods for forming generative models, for utilizing those models, or both. In embodiments, an object model fitting system can be developed comprising a 3D active appearance model (AAM) model. The 3D AAM comprises an appearance model comprising a set of subcomponent appearance models that is constrained by a 3D shape model. In embodiments, the 3D AAM may be generated using a balanced set of training images. The object model fitting system may further comprise one or more manifold constraints, one or more weighting factors, or both. Applications of the present invention include, but are not limited to, modeling and/or fitting face images, although the teachings of the present invention can be applied to modeling/fitting other objects.

US Patent:

8260038, Sep 4, 2012

Filed:

Feb 25, 2009

Appl. No.:

12/392840

Inventors:

Jing Xiao - Cupertino CA, US
Derek Shiell - Palo Alto CA, US

Assignee:

Seiko Epson Corporation - Tokyo

International Classification:

G06K 9/00
G06T 17/00

US Classification:

382154, 345420

Abstract:

Aspects of the present invention include systems and methods for forming generative models, for utilizing those models, or both. In embodiments, an object model fitting system can be developed comprising a 3D active appearance model (AAM) model. The 3D AAM comprises an appearance model comprising a set of subcomponent appearance models that is constrained by a 3D shape model. In embodiments, the 3D AAM may be generated using a balanced set of training images. The object model fitting system may further comprise one or more manifold constraints, one or more weighting factors, or both. Applications of the present invention include, but are not limited to, modeling and/or fitting face images, although the teachings of the present invention can be applied to modeling/fitting other objects.

US Patent:

8306257, Nov 6, 2012

Filed:

Jan 31, 2011

Appl. No.:

13/017891

Inventors:

Derek Shiell - Los Angeles CA, US
Jing Xiao - Cupertino CA, US

Assignee:

Seiko Epson Corporation - Tokyo

International Classification:

G06K 9/00
H04N 5/225

US Classification:

382100, 348169

Abstract:

An active appearance model is built by arranging the training images in its training library into a hierarchical tree with the training images at each parent node being divided into two child nodes according to similarities in characteristic features. The number of node levels is such that the number of training images associated with each leaf node is smaller than a predefined maximum. A separate AAM, one per leaf node, is constructed using each leaf node's corresponding training images. In operation, starting at the root node, a test image is compared with each parent node's two child nodes and follows a node-path of model images that most closely matches the test image. The test image is submitted to an AAM selected for being associated with the leaf node at which the test image rests. The selected AAM's output aligned image may be resubmitted to the hierarchical tree if sufficient alignment is not achieved.

US Patent:

8311319, Nov 13, 2012

Filed:

Dec 6, 2010

Appl. No.:

12/961347

Inventors:

Derek Shiell - Los Angeles CA, US
Jing Xiao - Cupertino CA, US

Assignee:

Seiko Epson Corporation - Tokyo

International Classification:

G06K 9/00
G06T 17/00
G06F 15/18

US Classification:

382159, 345420, 706 12

Abstract:

An Active Appearance Model, AAM, uses an Lminimization-based approach to aligning an input test image. In each iterative application of its statistical model fitting function, a shape parameter coefficient p and an appearance parameter coefficient λ within the statistical model fitting function are updated by Lminimization. The AAM further includes a canonical classifier to determine if an aligned image is a true example of the class of object being sought before the AAM is permitted to output its aligned image.

US Patent:

8208717, Jun 26, 2012

Filed:

Feb 25, 2009

Appl. No.:

12/392828

Inventors:

Jing Xiao - Cupertino CA, US
Derek Shiell - Palo Alto CA, US

Assignee:

Seiko Epson Corporation - Tokyo

International Classification:

G06K 9/00
G06T 17/00
G06F 19/00

US Classification:

382154, 345420, 700 98

Abstract:

Aspects of the present invention include systems and methods for forming generative models, for utilizing those models, or both. In embodiments, an object model fitting system can be developed comprising a 3D active appearance model (AAM) model. The 3D AAM comprises an appearance model comprising a set of subcomponent appearance models that is constrained by a 3D shape model. In embodiments, the 3D AAM may be generated using a balanced set of training images. The object model fitting system may further comprise one or more manifold constraints, one or more weighting factors, or both. Applications of the present invention include, but are not limited to, modeling and/or fitting face images, although the teachings of the present invention can be applied to modeling/fitting other objects.

US Patent:

20100013832, Jan 21, 2010

Filed:

Feb 25, 2009

Appl. No.:

12/392808

Inventors:

Jing Xiao - Cupertino CA, US
Derek Shiell - Palo Alto CA, US

International Classification:

G06T 17/00

US Classification:

345420, 382159, 382100

Abstract:

Aspects of the present invention include systems and methods for forming generative models, for utilizing those models, or both. In embodiments, an object model fitting system can be developed comprising a 3D active appearance model (AAM) model. The 3D AAM comprises an appearance model comprising a set of subcomponent appearance models that is constrained by a 3D shape model. In embodiments, the 3D AAM may be generated using a balanced set of training images. The object model fitting system may further comprise one or more manifold constraints, one or more weighting factors, or both. Applications of the present invention include, but are not limited to, modeling and/or fitting face images, although the teachings of the present invention can be applied to modeling/fitting other objects.

US Patent:

20120299906, Nov 29, 2012

Filed:

May 24, 2011

Appl. No.:

13/114514

Inventors:

Derek Shiell - Los Angeles CA, US
Jing Xiao - Cupertino CA, US

International Classification:

G06T 15/00

US Classification:

345419

Abstract:

An output image of higher resolution than an input image is constructed by using a low resolution (LR) dictionary of triangle data entries, each having a one-to-one correlation with a high resolution (HR) data entry in an HR dictionary of triangle data entries. The input image is triangularized, and the closest matching LR data entry in the LR dictionary for each triangle in the triangularized input image is identified. The HR data entry correlated to each identified matching LR data entry is then used to construct the output image by wrapping the correlated HR data entry onto the corresponding triangle on the triangularized input image.