公开(公告)号：US20240330361A1

公开(公告)日：2024-10-03

申请号：US18741082

申请日：2024-06-12

Applicant: Google LLC

Inventor： Zhen Li ,  Yi-Ting Chen ,  Yaxi Gao ,  Da-Cheng Juan ,  Aleksei Timofeev ,  Chun-Ta Lu ,  Futang Peng ,  Sujith Ravi ,  Andrew Tomkins ,  Thomas J. Duerig

IPC: G06F16/55 ,  G06F16/538 ,  G06F16/9538 ,  G06F18/214 ,  G06F18/22 ,  G06F18/40 ,  G06N3/042 ,  G06N3/044 ,  G06N3/084

CPC classification number: G06F16/55 ,  G06F16/538 ,  G06F16/9538 ,  G06F18/2148 ,  G06F18/22 ,  G06F18/41 ,  G06N3/042 ,  G06N3/044 ,  G06N3/084

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an image embedding model. In one aspect, a method comprises: obtaining training data comprising a plurality of training examples, wherein each training example comprises: an image pair comprising a first image and a second image; and selection data indicating one or more of: (i) a co-click rate of the image pair, and (ii) a similar-image click rate of the image pair; and using the training data to train an image embedding model having a plurality of image embedding model parameters.

公开(公告)号：US20230205813A1

公开(公告)日：2023-06-29

申请号：US18171511

申请日：2023-02-20

Applicant: Google LLC

Inventor： Zhen Li ,  Yi-Ting Chen ,  Yaxi Gao ,  Da-Cheng Juan ,  Aleksei Timofeev ,  Chun-Ta Lu ,  Futang Peng ,  Sujith Ravi ,  Andrew Tomkins ,  Thomas J. Duerig

IPC: G06F16/55 ,  G06F16/538 ,  G06F16/9538 ,  G06N3/084 ,  G06F18/22 ,  G06F18/40 ,  G06F18/214 ,  G06N3/042 ,  G06N3/044

CPC classification number: G06F16/55 ,  G06F16/538 ,  G06F16/9538 ,  G06N3/084 ,  G06F18/22 ,  G06F18/41 ,  G06F18/2148 ,  G06N3/042 ,  G06N3/044

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an image embedding model. In one aspect, a method comprises: obtaining training data comprising a plurality of training examples, wherein each training example comprises: an image pair comprising a first image and a second image; and selection data indicating one or more of: (i) a co-click rate of the image pair, and (ii) a similar-image click rate of the image pair; and using the training data to train an image embedding model having a plurality of image embedding model parameters.

公开(公告)号：US12038970B2

公开(公告)日：2024-07-16

申请号：US18171511

申请日：2023-02-20

Applicant: Google LLC

Inventor： Zhen Li ,  Yi-Ting Chen ,  Yaxi Gao ,  Da-Cheng Juan ,  Aleksei Timofeev ,  Chun-Ta Lu ,  Futang Peng ,  Sujith Ravi ,  Andrew Tomkins ,  Thomas J. Duerig

IPC: G06F16/00 ,  G06F16/538 ,  G06F16/55 ,  G06F16/9538 ,  G06F18/214 ,  G06F18/22 ,  G06F18/40 ,  G06N3/042 ,  G06N3/044 ,  G06N3/084

CPC classification number: G06F16/55 ,  G06F16/538 ,  G06F16/9538 ,  G06F18/2148 ,  G06F18/22 ,  G06F18/41 ,  G06N3/042 ,  G06N3/044 ,  G06N3/084

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an image embedding model. In one aspect, a method comprises: obtaining training data comprising a plurality of training examples, wherein each training example comprises: an image pair comprising a first image and a second image; and selection data indicating one or more of: (i) a co-click rate of the image pair, and (ii) a similar-image click rate of the image pair; and using the training data to train an image embedding model having a plurality of image embedding model parameters.