公开(公告)号：US20140376804A1

公开(公告)日：2014-12-25

申请号：US13923639

申请日：2013-06-21

Applicant: Xerox Corporation

Inventor： Zeynep Akata ,  Florent C. Perronnin ,  Zaid Harchaoui ,  Cordelia L. Schmid

IPC: G06K9/62 ,  G06K9/66

CPC classification number: G06K9/6267 ,  G06K9/00496 ,  G06K9/4633 ,  G06K9/4676 ,  G06K9/4685 ,  G06K9/6232 ,  G06K9/6234 ,  G06K9/6244 ,  G06K9/628

Abstract: In image classification, each class of a set of classes is embedded in an attribute space where each dimension of the attribute space corresponds to a class attribute. The embedding generates a class attribute vector for each class of the set of classes. A set of parameters of a prediction function operating in the attribute space respective to a set of training images annotated with classes of the set of classes is optimized such that the prediction function with the optimized set of parameters optimally predicts the annotated classes for the set of training images. The prediction function with the optimized set of parameters is applied to an input image to generate at least one class label for the input image. The image classification does not include applying a class attribute classifier to the input image.

Abstract translation: 在图像分类中，一组类的每个类嵌入在属性空间中，其中属性空间的每个维对应于一个类属性。 嵌入生成一组类的每个类的类属性向量。 优化在属性空间中操作的预测函数的一组参数，该属性空间分别对应于该类类的类别注释的一组训练图像，使得具有优化参数集合的预测函数最优地预测该组的注释类 训练图像。 将具有优化参数集的预测函数应用于输入图像以生成用于输入图像的至少一个类标签。 图像分类不包括将类属性分类器应用于输入图像。

公开(公告)号：US10331976B2

公开(公告)日：2019-06-25

申请号：US13923639

申请日：2013-06-21

Applicant: Xerox Corporation

Inventor： Zeynep Akata ,  Florent C. Perronnin ,  Zaid Harchaoui ,  Cordelia L. Schmid

IPC: G06K9/62 ,  G06K9/46 ,  G06K9/00

Abstract: In image classification, each class of a set of classes is embedded in an attribute space where each dimension of the attribute space corresponds to a class attribute. The embedding generates a class attribute vector for each class of the set of classes. A set of parameters of a prediction function operating in the attribute space respective to a set of training images annotated with classes of the set of classes is optimized such that the prediction function with the optimized set of parameters optimally predicts the annotated classes for the set of training images. The prediction function with the optimized set of parameters is applied to an input image to generate at least one class label for the input image. The image classification does not include applying a class attribute classifier to the input image.