公开(公告)号：US20160140438A1

公开(公告)日：2016-05-19

申请号：US14884600

申请日：2015-10-15

Applicant: NEC Laboratories America, Inc.

Inventor： Tianbao Yang ,  Xiaoyu Wang ,  Yuanqing Lin ,  Saining Xie

IPC: G06N3/08

CPC classification number: G06N3/08 ,  G06N3/0454 ,  G06N3/084

Abstract: Systems and methods are disclosed for training a learning machine by augmenting data from fine-grained image recognition with labeled data annotated by one or more hyper-classes, performing multi-task deep learning; allowing fine-grained classification and hyper-class classification to share and learn the same feature layers; and applying regularization in the multi-task deep learning to exploit one or more relationships between the fine-grained classes and the hyper-classes.

Abstract translation: 公开了用于训练学习机器的系统和方法，通过用由一个或多个超类注释的标记数据增强来自细粒度图像识别的数据，执行多任务深度学习; 允许细粒度分类和超类分类来共享和学习相同的特征层; 并在多任务深度学习中应用正则化，以利用细粒度类与超类之间的一个或多个关系。