DETECTING OBJECTS IN IMAGES BY GENERATING SEQUENCES OF TOKENS

    公开(公告)号:US20250139959A1

    公开(公告)日:2025-05-01

    申请号:US18690550

    申请日:2022-09-19

    Applicant: Google LLC

    Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for object detection using neural networks. In one aspect, one of the methods includes obtaining an input image; processing the input image using an object detection neural network to generate an output sequence that comprises respective token at each of a plurality of time steps, wherein each token is selected from a vocabulary of tokens that comprises (i) a first set of tokens that each represent a respective discrete number from a set of discretized numbers and (ii) a second set of tokens that each represent a respective object category from a set of object categories; and generating, from the tokens in the output sequence, an object detection output for the input image.

    Modeling lift of metrics for triggering push notifications

    公开(公告)号:US11574339B1

    公开(公告)日:2023-02-07

    申请号:US16705919

    申请日:2019-12-06

    Applicant: Google LLC

    Abstract: Processor(s) of a client device can: analyze one or more features of an electronic resource that is under consideration for solicitation to a user; determine a notification likelihood that the user will access the electronic resource in response to an unsolicited notification of the electronic resource being output to the user; determine a baseline likelihood that the user will access the electronic resource without being solicited; compare the notification likelihood with the baseline likelihood; and cause, based on the comparing, the unsolicited notification to be output to the user. In some implementations, determining the notification likelihood and/or the baseline likelihood is based on applying data associated with the electronic resource as input across a machine learning model to generate output indicative of the notification likelihood and/or the baseline likelihood. In other implementations, determining the notification likelihood and/or the baseline likelihood is based on past behavior or preference(s) of the user.

Patent Agency Ranking