公开(公告)号：US12198420B2

公开(公告)日：2025-01-14

申请号：US17727100

申请日：2022-04-22

Applicant: eBay Inc.

Inventor： Kevin Lin ,  Fan Yang ,  Qiaosong Wang ,  Robinson Piramuthu

IPC: G06N3/088 ,  G06F16/51 ,  G06F16/532 ,  G06F16/56 ,  G06F16/583 ,  G06F18/21 ,  G06F18/214 ,  G06F18/22 ,  G06F18/2411 ,  G06V10/44 ,  G06V10/82 ,  G06V20/00

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for using adversarial learning for fine-grained image search. An image search system receives a search query that includes an input image depicting an object. The search system generates, using a generator, a vector representation of the object in a normalized view. The generator was trained based on a set of reference images of known objects in multiple views, and feedback data received from an evaluator that indicates performance of the generator at generating vector representations of the known objects in the normalized view. The evaluator including a discriminator sub-module, a normalizer sub-module, and a semantic embedding sub-module that generate the feedback data. The image search system identifies, based on the vector representation of the object, a set of other images depicting the object, and returns at least one of the other images in response to the search query.

公开(公告)号：US11361191B2

公开(公告)日：2022-06-14

申请号：US15985818

申请日：2018-05-22

Applicant: eBay Inc.

Inventor： Kevin Lin ,  Fan Yang ,  Qiaosong Wang ,  Robinson Piramuthu

IPC: G06K9/62 ,  G06F16/51 ,  G06F16/532 ,  G06F16/56 ,  G06N3/08 ,  G06V20/00 ,  G06F16/583 ,  G06V10/32

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for using adversarial learning for fine-grained image search. An image search system receives a search query that includes an input image depicting an object. The search system generates, using a generator, a vector representation of the object in a normalized view. The generator was trained based on a set of reference images of known objects in multiple views, and feedback data received from an evaluator that indicates performance of the generator at generating vector representations of the known objects in the normalized view. The evaluator including a discriminator sub-module, a normalizer sub-module, and a semantic embedding sub-module that generate the feedback data. The image search system identifies, based on the vector representation of the object, a set of other images depicting the object, and returns at least one of the other images in response to the search query.

公开(公告)号：US20220245406A1

公开(公告)日：2022-08-04

申请号：US17727100

申请日：2022-04-22

Applicant: eBay Inc.

Inventor： Kevin Lin ,  Fan Yang ,  Qiaosong Wang ,  Robinson Piramuthu

IPC: G06K9/62 ,  G06F16/583 ,  G06F16/51 ,  G06F16/532 ,  G06F16/56 ,  G06N3/08 ,  G06V10/32

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for using adversarial learning for fine-grained image search. An image search system receives a search query that includes an input image depicting an object. The search system generates, using a generator, a vector representation of the object in a normalized view. The generator was trained based on a set of reference images of known objects in multiple views, and feedback data received from an evaluator that indicates performance of the generator at generating vector representations of the known objects in the normalized view. The evaluator including a discriminator sub-module, a normalizer sub-module, and a semantic embedding sub-module that generate the feedback data. The image search system identifies, based on the vector representation of the object, a set of other images depicting the object, and returns at least one of the other images in response to the search query.