公开(公告)号：US20220108131A1

公开(公告)日：2022-04-07

申请号：US17062157

申请日：2020-10-02

Applicant: Adobe Inc.

Inventor： Jason Wen Yong Kuen ,  Zhe Lin ,  Jiuxiang Gu

IPC: G06K9/62 ,  G06N3/04 ,  G06T3/40 ,  G06T3/60

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for accurately and efficiently learning parameters of a distilled neural network from parameters of a source neural network utilizing multiple augmentation strategies. For example, the disclosed systems can generate lightly augmented digital images and heavily augmented digital images. The disclosed systems can further learn parameters for a source neural network from the lightly augmented digital images. Moreover, the disclosed systems can learn parameters for a distilled neural network from the parameters learned for the source neural network. For example, the disclosed systems can compare classifications of heavily augmented digital images generated by the source neural network and the distilled neural network to transfer learned parameters from the source neural network to the distilled neural network via a knowledge distillation loss function.

公开(公告)号：US20200272822A1

公开(公告)日：2020-08-27

申请号：US16874114

申请日：2020-05-14

Applicant: Adobe Inc.

Inventor： Zhe Lin ,  Xiaohui Shen ,  Mingyang Ling ,  Jianming Zhang ,  Jason Wen Yong Kuen

IPC: G06K9/00 ,  G06K9/62 ,  G06N3/04

Abstract: In implementations of object detection in images, object detectors are trained using heterogeneous training datasets. A first training dataset is used to train an image tagging network to determine an attention map of an input image for a target concept. A second training dataset is used to train a conditional detection network that accepts as conditional inputs the attention map and a word embedding of the target concept. Despite the conditional detection network being trained with a training dataset having a small number of seen classes (e.g., classes in a training dataset), it generalizes to novel, unseen classes by concept conditioning, since the target concept propagates through the conditional detection network via the conditional inputs, thus influencing classification and region proposal. Hence, classes of objects that can be detected are expanded, without the need to scale training databases to include additional classes.