知搜-全球专利检索

  1. Login
  2. Sign Up

Search Results

79 records (took 0.08 seconds)

    COMMUNICATION EFFICIENT MACHINE LEARNING OF DATA ACROSS MULTIPLE SITES
    2.
    发明申请
    COMMUNICATION EFFICIENT MACHINE LEARNING OF DATA ACROSS MULTIPLE SITES 审中-公开

    公开(公告)号：US20200090002A1

    公开(公告)日：2020-03-19

    申请号：US16131150

    申请日：2018-09-14

    Applicant: Cisco Technology, Inc.

    Inventor： Xiaoqing Zhu ,  Yaqi Wang ,  Dan Tan ,  Rob Liston ,  Mehdi Nikkhah

    IPC: G06K9/62 ,  G06F15/18 ,  G06N3/04 ,  H04Q9/00

    Abstract: In one embodiment, a service receives machine learning-based generative models from a plurality of distributed sites. Each generative model is trained locally at a site using unlabeled data observed at that site to generate synthetic unlabeled data that mimics the unlabeled data used to train the generative model. The service receives, from each of the distributed sites, a subset of labeled data observed at that site. The service uses the generative models to generate synthetic unlabeled data. The service trains a global machine learning-based model using the received subsets of labeled data received from the distributed sites and the synthetic unlabeled data generated by the generative models.

    Clustering-based person re-identification
    4.
    发明授权
    Clustering-based person re-identification 有权

    公开(公告)号：US10275683B2

    公开(公告)日：2019-04-30

    申请号：US15409821

    申请日：2017-01-19

    Applicant: Cisco Technology, Inc.

    Inventor： Xiaoqing Zhu ,  Rob Liston ,  John G. Apostolopoulos ,  Wai-tian Tan

    IPC: G06K9/00 ,  G06K9/62 ,  G06F17/30

    Abstract: Presented herein are techniques for assignment of an identity to a group of captured images. A plurality of captured images that each include an image of at least one person are obtained. For each of the plurality of captured images, relational metrics indicating a relationship between the image of the person in a respective captured image and the images of the persons in each of the remaining plurality of captured images is calculated. Based on the relational metrics, a clustering process is performed to generate one or more clusters from the plurality of captured images. Each of the one or more clusters are associated with an identity of an identity database. The one or more clusters may each be associated with an existing identity of the identity database or an additional identity that is not yet present in the identity database.

    DYNAMIC PROGRAMMING ACROSS MULTIPLE STREAMS
    8.
    发明申请
    DYNAMIC PROGRAMMING ACROSS MULTIPLE STREAMS 有权

    公开(公告)号：US20180035140A1

    公开(公告)日：2018-02-01

    申请号：US15728681

    申请日：2017-10-10

    Applicant: Cisco Technology, Inc.

    Inventor： Zhi Li ,  Joshua B. Gahm ,  Xiaoqing Zhu

    IPC: H04N21/2343 ,  H04N21/647 ,  H04N21/2385 ,  H04N21/238 ,  H04L29/06 ,  H04N21/845 ,  H04N21/2662 ,  H04L12/825

    CPC classification number: H04N21/234381 ,  H04L47/25 ,  H04L65/602 ,  H04L65/605 ,  H04L65/80 ,  H04N21/23439 ,  H04N21/23805 ,  H04N21/2385 ,  H04N21/2662 ,  H04N21/64769 ,  H04N21/8456

    Abstract: Various implementations disclosed herein enable a more efficient allocation of one or more shared network resources (e.g., bandwidth, memory, processor time, etc.) amongst a number of client devices based on media content data complexity and client device resource constraints in order to better manage perceptual playback quality of adaptive streaming content. In some implementations, a method includes aligning sequences of one or more temporal segments such that time boundaries of temporal segments across the sequences are in alignment; and, selecting segment representations for each temporal segment based on a combination of the sequence alignment and perceptual quality level values associated with available segment representations, such that a combination of resulting perceptual quality levels satisfies a joint quality criterion. Each sequence is associated with a respective one of a number of client devices sharing a network resource and an instance of a respective video stream. The one or more temporal segments of each sequence are used to provide segment representations of media content data to one of the client devices. The alignment of time boundaries of temporal segments is achieved at least in part by adjusting performance characteristics associated with at least some of the one or more temporal segments.

    Augmenting Wi-Fi Localization with Auxiliary Sensor Information
    10.
    发明申请
    Augmenting Wi-Fi Localization with Auxiliary Sensor Information 审中-公开

    公开(公告)号：US20170188194A1

    公开(公告)日：2017-06-29

    申请号：US15397407

    申请日：2017-01-03

    Applicant: Cisco Technology, Inc.

    Inventor： John George Apostolopoulos ,  Wai-tian Tan ,  Xiaoqing Zhu

    IPC: H04W4/02

    CPC classification number: H04W4/023 ,  H04L61/6022 ,  H04W4/025 ,  H04W8/26 ,  H04W84/12

    Abstract: In one implementation, a method of maintaining continuous identity for mobile devices includes: obtaining a first address for a first device; and obtaining, from one or more auxiliary sensors, auxiliary sensor information related to the first device. The method also includes determining whether the auxiliary sensor information matches information associated with a second address, where the second address was previously associated with the first device. The method further includes linking the first address with the second address for the first device, in order to continue tracking the first device when the second address is no longer detected, in response to determining that the auxiliary sensor information matches information associated with the second address.

Patent Agency Ranking