Patent search ap:("Google LLC") AND inv:"Ting Liu" Page 1

1.

发明申请
MEDIA TREND DETECTION AND MAINTENANCE AT A CONTENT SHARING PLATFORM 有权

公开(公告)号：US20250111675A1

公开(公告)日：2025-04-03

申请号：US18900467

申请日：2024-09-27

Applicant: Google LLC

Inventor： Hui Miao , Chun-Te Chu , Mingyan Gao , Huanfen Yao , Ting Liu , Long Zhao , Liangzhe Yuan , Yukun Zhu , Vinay Kumar Bettadapura , Ye Jin

IPC: G06V20/40 , G06V10/74 , G06V10/75 , G06V10/762 , G06V10/80

Abstract: Methods and systems for media trend detection and maintenance are provided herein. A set of media items each having common media characteristics is identified. A set of pose values is determined for each respective media item of the set of media items. Each pose value is associated with a particular predefined pose for objects depicted by the set of media items. A set of distance scores is calculated. Each distance score represents a distance between the respective set of pose values determined for a media item and a respective set of pose values determined for an additional media item. A coherence score is determined for the set of media items based on the calculated set of distance scores. Responsive to a determination that the coherence score satisfies one or more coherence criteria, a determination is made that the set of media items corresponds to a media trend of a platform.

2.

发明授权
Weakly-supervised action localization by sparse temporal pooling network 有权

公开(公告)号：US11881022B2

公开(公告)日：2024-01-23

申请号：US18181806

申请日：2023-03-10

Applicant: Google LLC

Inventor： Ting Liu , Gautam Prasad , Phuc Xuan Nguyen , Bohyung Han

IPC: G06V20/40 , G06F18/214 , G06F18/243

CPC classification number: G06V20/40 , G06F18/214 , G06F18/24317 , G06V20/44

Abstract: Systems and methods for a weakly supervised action localization model are provided. Example models according to example aspects of the present disclosure can localize and/or classify actions in untrimmed videos using machine-learned models, such as convolutional neural networks. The example models can predict temporal intervals of human actions given video-level class labels with no requirement of temporal localization information of actions. The example models can recognize actions and identify a sparse set of keyframes associated with actions through adaptive temporal pooling of video frames, wherein the loss function of the model is composed of a classification error and a sparsity of frame selection. Following action recognition with sparse keyframe attention, temporal proposals for action can be extracted using temporal class activation mappings, and final time intervals can be estimated corresponding to target actions.

3.

发明授权
Providing online content 有权

公开(公告)号：US11798009B1

公开(公告)日：2023-10-24

申请号：US17567439

申请日：2022-01-03

Applicant: Google LLC

Inventor： Ting Liu , Zhengzhu Feng , Zhongyi Lin

IPC: G06Q30/02

CPC classification number: G06Q30/02

Abstract: Systems and methods for providing online content include evaluating a custom selection rule specified by a content provider. The custom selection rule may be used to control whether content from the provider is eligible for selection by a content selection service. The content selection rule may include one or more logical operators, a selected interest category and/or a selected list of one or more client identifiers.

4.

发明授权
Weakly-supervised action localization by sparse temporal pooling network 有权

公开(公告)号：US11640710B2

公开(公告)日：2023-05-02

申请号：US16625172

申请日：2018-11-05

Applicant: Google LLC

Inventor： Ting Liu , Gautam Prasad , Phuc Xuan Nguyen , Bohyung Han

IPC: G06V20/40 , G06K9/62

Abstract: Systems and methods for a weakly supervised action localization model are provided. Example models according to example aspects of the present disclosure can localize and/or classify actions in untrimmed videos using machine-learned models, such as convolutional neural networks. The example models can predict temporal intervals of human actions given video-level class labels with no requirement of temporal localization information of actions. The example models can recognize actions and identify a sparse set of keyframes associated with actions through adaptive temporal pooling of video frames, wherein the loss function of the model is composed of a classification error and a sparsity of frame selection. Following action recognition with sparse keyframe attention, temporal proposals for action can be extracted using temporal class activation mappings, and final time intervals can be estimated corresponding to target actions.

5.

发明授权
Providing online content 有权

公开(公告)号：US11216829B1

公开(公告)日：2022-01-04

申请号：US16283905

申请日：2019-02-25

Applicant: Google LLC

Inventor： Ting Liu , Zhengzhu Feng , Zhongyi Lin

IPC: G06Q30/02

Abstract: Systems and methods for providing online content include evaluating a custom selection rule specified by a content provider. The custom selection rule may be used to control whether content from the provider is eligible for selection by a content selection service. The content selection rule may include one or more logical operators, a selected interest category and/or a selected list of one or more client identifiers.

6.

发明申请
Weakly-Supervised Action Localization by Sparse Temporal Pooling Network 审中-公开

公开(公告)号：US20200272823A1

公开(公告)日：2020-08-27

申请号：US16625172

申请日：2018-11-05

Applicant: Google LLC

Inventor： Ting Liu , Gautam Prasad , Phuc Xuan Nguyen , Bohyung Han

IPC: G06K9/00 , G06K9/62

Abstract: Systems and methods for a weakly supervised action localization model are provided. Example models according to example aspects of the present disclosure can localize and/or classify actions in untrimmed videos using machine-learned models, such as convolutional neural networks. The example models can predict temporal intervals of human actions given video-level class labels with no requirement of temporal localization information of actions. The example models can recognize actions and identify a sparse set of keyframes associated with actions through adaptive temporal pooling of video frames, wherein the loss function of the model is composed of a classification error and a sparsity of frame selection. Following action recognition with sparse keyframe attention, temporal proposals for action can be extracted using temporal class activation mappings, and final time intervals can be estimated corresponding to target actions.

7.

发明公开
Weakly-Supervised Action Localization by Sparse Temporal Pooling Network 审中-公开

公开(公告)号：US20230215169A1

公开(公告)日：2023-07-06

申请号：US18181806

申请日：2023-03-10

Applicant: Google LLC

Inventor： Ting Liu , Gautam Prasad , Phuc Xuan Nguyen , Bohyung Han

IPC: G06V20/40 , G06F18/214 , G06F18/243

CPC classification number: G06V20/40 , G06F18/214 , G06F18/24317 , G06V20/44

Abstract: Systems and methods for a weakly supervised action localization model are provided. Example models according to example aspects of the present disclosure can localize and/or classify actions in untrimmed videos using machine-learned models, such as convolutional neural networks. The example models can predict temporal intervals of human actions given video-level class labels with no requirement of temporal localization information of actions. The example models can recognize actions and identify a sparse set of keyframes associated with actions through adaptive temporal pooling of video frames, wherein the loss function of the model is composed of a classification error and a sparsity of frame selection. Following action recognition with sparse keyframe attention, temporal proposals for action can be extracted using temporal class activation mappings, and final time intervals can be estimated corresponding to target actions.