-
公开(公告)号：US10719509B2
公开(公告)日：2020-07-21
申请号：US15290198
申请日：2016-10-11
Applicant: GOOGLE INC.
Inventor： Sanjiv Kumar , David Morris Simcha , Ananda Theertha Suresh , Ruiqi Guo , Xinnan Yu , Daniel Holtmann-Rice
IPC: G06F16/2453 , G06F16/28 , G06F16/22 , G06F16/2457 , G06F17/10 , G06K9/62 , G06F16/33 , G06F16/35
Abstract: Implementations provide an efficient system for calculating inner products between high-dimensionality vectors. An example method includes clustering database items represented as vectors, selecting a cluster center for each cluster, and storing the cluster center as an entry in a first layer codebook. The method also includes, for each database item, calculating a residual based on the cluster center for the cluster the database item is assigned to and projecting the residual into subspaces. The method also includes determining, for each of the subspaces, an entry in a second layer codebook for the subspace, and storing the entry in the first layer codebook and the respective entry in the second layer codebook for each of the subspaces as a quantized vector for the database item. The entry can be used to categorize an item represented by a query vector or to provide database items responsive to a query vector.
-
公开(公告)号：US20160335053A1
公开(公告)日：2016-11-17
申请号：US14710467
申请日：2015-05-12
Applicant: Google Inc.
Inventor： Sanjiv Kumar , Xinnan Yu
CPC classification number: G06F5/017 , G06K9/6232 , G06N99/005 , H03M7/30
Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for augmenting neural networks with an external memory. One of the methods includes receiving a plurality of high-dimensional data items; generating a circulant embedding matrix for the high-dimensional data items, wherein the circulant embedding matrix is a matrix that is fully specified by a single vector; for each high-dimensional data item, generating a compact representation of the high-dimensional data item, comprising computing a product of the circulant embedding matrix and the high dimensional data item by performing a circular convolution of the single vector that fully specifies the circulant embedding matrix and the high dimensional data item using a Fast Fourier Transform (FFT); and generating a compact representation of the high dimensional data item by computing a binary map of the computed product.
Abstract translation: 方法，系统和装置，包括在计算机存储介质上编码的计算机程序，用于利用外部存储器增强神经网络。 方法之一包括接收多个高维数据项; 为高维数据项生成循环嵌入矩阵，其中循环嵌入矩阵是由单个向量完全指定的矩阵; 对于每个高维数据项，生成高维数据项的紧凑表示，包括通过执行完全指定循环嵌入的单个向量的循环卷积来计算循环嵌入矩阵和高维数据项的乘积 矩阵和使用快速傅立叶变换（FFT）的高维数据项; 以及通过计算所述计算产品的二进制图来生成所述高维数据项的紧凑表示。
-
公开(公告)号：US20180089587A1
公开(公告)日：2018-03-29
申请号：US15676076
申请日：2017-08-14
Applicant: Google Inc.
Inventor： Ananda Theertha Suresh , Sanjiv Kumar , Hugh Brendan McMahan , Xinnan Yu
Abstract: The present disclosure provides systems and methods for communication efficient distributed mean estimation. In particular, aspects of the present disclosure can be implemented by a system in which a number of vectors reside on a number of different clients, and a centralized server device seeks to estimate the mean of such vectors. According to one aspect of the present disclosure, a client computing device can rotate a vector by a random rotation matrix and then subsequently perform probabilistic quantization on the rotated vector. According to another aspect of the present disclosure, subsequent to quantization but prior to transmission, the client computing can encode the quantized vector according to a variable length coding scheme (e.g., by computing variable length codes).
-
Patent Agency Ranking
公开(公告)号：US11196800B2
公开(公告)日：2021-12-07
申请号：US15708793
申请日：2017-09-19
Applicant: Google Inc.
Inventor： Ananda Theertha Suresh , Sanjiv Kumar , Hugh Brendan McMahan , Xinnan Yu
IPC: H04L29/08 , G06N7/00 , G06F17/16 , H03M7/30 , H03M7/40 , G06F17/12 , G06N20/00 , G06F17/18 , H04L29/06
Abstract: The present disclosure provides systems and methods for communication efficient distributed mean estimation. In particular, aspects of the present disclosure can be implemented by a system in which a number of vectors reside on a number of different clients, and a centralized server device seeks to estimate the mean of such vectors. According to one aspect of the present disclosure, a client computing device can rotate a vector by a random rotation matrix and then subsequently perform probabilistic quantization on the rotated vector. According to another aspect of the present disclosure, subsequent to quantization but prior to transmission, the client computing can encode the quantized vector according to a variable length coding scheme (e.g., by computing variable length codes).
-
公开(公告)号：US20190294967A1
公开(公告)日：2019-09-26
申请号：US15014804
申请日：2016-02-03
Applicant: Google Inc.
Inventor： Sanjiv Kumar , Xinnan Yu
Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing inputs using a neural network that includes a circulant neural network layer. One of the methods includes receiving a layer input for the circulant layer; and processing the layer input to generate a layer output for the circulant layer, wherein processing the layer input comprises computing an activation function, wherein the activation function is dependent on the product of the circulant matrix associated with the circulant layer and the layer input, and wherein computing the activation function comprises performing a circular convolution using a Fast Fourier Transform (FFT).
-
公开(公告)号：US20180089590A1
公开(公告)日：2018-03-29
申请号：US15708793
申请日：2017-09-19
Applicant: Google Inc.
Inventor： Ananda Theertha Suresh , Sanjiv Kumar , Hugh Brendan McMahan , Xinnan Yu
CPC classification number: G06N20/00 , G06F17/12 , G06F17/16 , G06F17/18 , G06N7/005 , H03M7/3059 , H03M7/3082 , H03M7/40 , H04L67/42
Abstract: The present disclosure provides systems and methods for communication efficient distributed mean estimation. In particular, aspects of the present disclosure can be implemented by a system in which a number of vectors reside on a number of different clients, and a centralized server device seeks to estimate the mean of such vectors. According to one aspect of the present disclosure, a client computing device can rotate a vector by a random rotation matrix and then subsequently perform probabilistic quantization on the rotated vector. According to another aspect of the present disclosure, subsequent to quantization but prior to transmission, the client computing can encode the quantized vector according to a variable length coding scheme (e.g., by computing variable length codes).
-
公开(公告)号：US09870199B2
公开(公告)日：2018-01-16
申请号：US14710467
申请日：2015-05-12
Applicant: Google Inc.
Inventor： Sanjiv Kumar , Xinnan Yu
CPC classification number: G06F5/017 , G06K9/6232 , G06N99/005 , H03M7/30
Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for augmenting neural networks with an external memory. One of the methods includes receiving a plurality of high-dimensional data items; generating a circulant embedding matrix for the high-dimensional data items, wherein the circulant embedding matrix is a matrix that is fully specified by a single vector; for each high-dimensional data item, generating a compact representation of the high-dimensional data item, comprising computing a product of the circulant embedding matrix and the high dimensional data item by performing a circular convolution of the single vector that fully specifies the circulant embedding matrix and the high dimensional data item using a Fast Fourier Transform (FFT); and generating a compact representation of the high dimensional data item by computing a binary map of the computed product.
-
公开(公告)号：US20170091240A1
公开(公告)日：2017-03-30
申请号：US14951909
申请日：2015-11-25
Applicant: Google Inc.
Inventor： Xinnan Yu , Sanjiv Kumar , Ruiqi Guo
IPC: G06F17/30
CPC classification number: G06F16/2237 , G06F16/3331 , G06F16/951
Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for efficiently performing linear projections. In one aspect, a method includes actions for obtaining a plurality of content items from one or more content sources. Additional actions include, extracting a plurality of features from each of the plurality of content items, generating a feature vector for each of the extracted features in order to create a search space, generating a series of element matrices based upon the generated feature vectors, transforming the series of element matrices into a structured matrix such that the transformation preserves one or more relationships associated with each element matrix of the series of element matrices, receiving a search object, searching the enhanced search space based on the received search object, provided one or more links to a content item that are responsive to the search object.
-
-
-
-
-
-
-
Search Results
8
records
(took 0.015 seconds)
