-
公开(公告)号：US20190065957A1
公开(公告)日：2019-02-28
申请号：US15690426
申请日：2017-08-30
申请人： Google Inc.
摘要： The present disclosure provides systems and methods that enable distance metric learning using proxies. A machine-learned distance model can be trained in a proxy space in which a loss function compares an embedding provided for an anchor data point of a training dataset to a positive proxy and one or more negative proxies, where each of the positive proxy and the one or more negative proxies serve as a proxy for two or more data points included in the training dataset. Thus, each proxy can approximate a number of data points, enabling faster convergence. According to another aspect, the proxies of the proxy space can themselves be learned parameters, such that the proxies and the model are trained jointly. Thus, the present disclosure enables faster convergence (e.g., reduced training time). The present disclosure provides example experiments which demonstrate a new state of the art on several popular training datasets.
-
公开(公告)号：US20190065899A1
公开(公告)日：2019-02-28
申请号：US15710377
申请日：2017-09-20
申请人： Google Inc.
CPC分类号： G06K9/6215 , G06K9/4628 , G06K9/6232 , G06K9/6255 , G06K9/6256 , G06K9/6262 , G06K9/66 , G06N20/00
摘要： The present disclosure provides systems and methods that enable distance metric learning using proxies. A machine-learned distance model can be trained in a proxy space in which a loss function compares an embedding provided for an anchor data point of a training dataset to a positive proxy and one or more negative proxies, where each of the positive proxy and the one or more negative proxies serve as a proxy for two or more data points included in the training dataset. Thus, each proxy can approximate a number of data points, enabling faster convergence. According to another aspect, the proxies of the proxy space can themselves be learned parameters, such that the proxies and the model are trained jointly. Thus, the present disclosure enables faster convergence (e.g., reduced training time). The present disclosure provides example experiments which demonstrate a new state of the art on several popular training datasets.
-
公开(公告)号：US09588990B1
公开(公告)日：2017-03-07
申请号：US14678507
申请日：2015-04-03
申请人： Google Inc.
发明人： Sergey Ioffe , Yushi Jing , Stephen C. Holiday
CPC分类号： G06F17/3025 , G06F17/30256 , G06K9/00684 , G06K9/62 , G06K9/6212 , G06K9/6262
摘要： Image similarity operations are performed in which a seed image is analyzed, and a set of semantic classifications are determined from analyzing the seed image. The set of semantic classifications can include multiple positive semantic classifications. A distance measure is determined that is specific to the set of semantic classifications. The seed image is compared to a collection of images using the distance measure. A set of similar images is determined from comparing the seed image to the collection of images.
摘要翻译： 执行图像相似度操作，其中分析种子图像，并且通过分析种子图像来确定一组语义分类。 语义分类集合可以包括多个正的语义分类。 确定特定于语义分类集合的距离度量。 使用距离测量将种子图像与图像的集合进行比较。 通过比较种子图像和图像的集合来确定一组相似的图像。
-
4.发明授权公开(公告)号：US09552549B1
公开(公告)日：2017-01-24
申请号：US14444272
申请日：2014-07-28
申请人： Google Inc.
IPC分类号： G06N3/08
CPC分类号： G06N3/084 , G06N3/0454
摘要： Systems and techniques are provided for a ranking approach to train deep neural nets for multilabel image annotation. Label scores may be received for labels determined by a neural network for training examples. Each label may be a positive label or a negative label for the training example. An error of the neural network may be determined based on a comparison, for each of the training examples, of the label scores for positive labels and negative labels for the training example and a semantic distance between each positive label and each negative label for the training example. Updated weights may be determined for the neural network based on a gradient of the determined error of the neural network. The updated weights may be applied to the neural network to train the neural network.
摘要翻译： 提供系统和技术用于排列方法来训练用于多标签图像注释的深层神经网络。 可以接收由用于训练示例的神经网络确定的标签的标签分数。 每个标签可能是培训示例的正标签或负标签。 可以基于针对训练样本的正标签的标签分数和训练样本的负标签的每个训练样本的比较以及训练样本的每个正标签和每个负标签之间的语义距离来确定神经网络的误差 例。 可以基于确定的神经网络的误差的梯度来确定神经网络的更新权重。 更新的权重可以应用于神经网络来训练神经网络。
-
公开(公告)号：US20160217368A1
公开(公告)日：2016-07-28
申请号：US15009647
申请日：2016-01-28
申请人： Google Inc.
发明人： Sergey Ioffe , Corinna Cortes
摘要： Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing inputs using a neural network system that includes a batch normalization layer. One of the methods includes receiving a respective first layer output for each training example in the batch; computing a plurality of normalization statistics for the batch from the first layer outputs; normalizing each component of each first layer output using the normalization statistics to generate a respective normalized layer output for each training example in the batch; generating a respective batch normalization layer output for each of the training examples from the normalized layer outputs; and providing the batch normalization layer output as an input to the second neural network layer.
摘要翻译： 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，用于使用包括批量归一化层的神经网络系统处理输入。 方法之一包括为批次中的每个训练样本接收相应的第一层输出; 从第一层输出计算批次的多个归一化统计; 使用归一化统计来对每个第一层输出的每个分量进行归一化，以为该批中的每个训练样本生成相应的归一化层输出; 从标准化层输出生成针对每个训练样本的相应批量标准化层输出; 以及提供所述批量归一化层输出作为第二神经网络层的输入。
-
公开(公告)号：US20150154743A1
公开(公告)日：2015-06-04
申请号：US14615029
申请日：2015-02-05
申请人： Google Inc.
发明人： Sergey Ioffe , Troy Chinen , Vivek Kwatra , Hui Fang , Yichang Shih
CPC分类号： G06T5/002 , G06K9/00221 , G06K9/00261 , G06K9/40 , G06T5/005 , G06T5/20 , G06T5/50 , G06T2207/10004 , G06T2207/10016 , G06T2207/20028 , G06T2207/20182 , G06T2207/20216 , G06T2207/30201
摘要： A method, computer program product, and computer system for identifying a first portion of a facial image in a first image, wherein the first portion includes noise. A corresponding portion of the facial image is identified in a second image, wherein the corresponding portion includes less noise than the first portion. One or more filter parameters of the first portion are determined based upon, at least in part, the first portion and the corresponding portion. At least a portion of the noise from the first portion is smoothed based upon, at least in part, the one or more filter parameters. At least a portion of face specific details from the corresponding portion is added to the first portion.
摘要翻译： 一种用于识别第一图像中的面部图像的第一部分的方法，计算机程序产品和计算机系统，其中所述第一部分包括噪声。 面部图像的相应部分在第二图像中被识别，其中对应部分包括比第一部分少的噪声。 至少部分地基于第一部分和对应部分来确定第一部分的一个或多个过滤器参数。 至少部分地基于一个或多个过滤器参数来平滑来自第一部分的噪声的至少一部分。 来自相应部分的面部特定细节的至少一部分被添加到第一部分。
-
公开(公告)号：US08965891B1
公开(公告)日：2015-02-24
申请号：US14083043
申请日：2013-11-18
申请人： Google Inc.
发明人： Samy Bengio , Gal Chechik , Sergey Ioffe , Jay Yagnik
CPC分类号： G06K9/66 , G06F17/30244 , G06F17/3053 , G06K9/6267 , Y10S707/913 , Y10S707/915
摘要： Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training scoring models. One method includes storing data identifying a plurality of positive and a plurality of negative training images for a query. The method further includes selecting a first image from either the positive group of images or the negative group of images, and applying a scoring model to the first image. The method further includes selecting a plurality of candidate images from the other group of images, applying the scoring model to each of the candidate images, and then selecting a second image from the candidate images according to scores for the images. The method further includes determining that the scores for the first image and the second image fail to satisfy a criterion, updating the scoring model, and storing the updated scoring model.
摘要翻译： 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，用于训练评分模型。 一种方法包括存储识别用于查询的多个正训练图像和多个负训练图像的数据。 该方法还包括从图像的正组或负图像组中选择第一图像，以及将评分模型应用于第一图像。 该方法还包括从另一组图像中选择多个候选图像，将评分模型应用于每个候选图像，然后根据图像的分数从候选图像中选择第二图像。 该方法还包括确定第一图像和第二图像的分数不能满足标准，更新评分模型，并存储更新的评分模型。
-
公开(公告)号：US20170132512A1
公开(公告)日：2017-05-11
申请号：US15343458
申请日：2016-11-04
申请人： Google Inc.
发明人： Sergey Ioffe
CPC分类号： G06N3/08 , G06K9/4628 , G06K9/6256 , G06K9/6272 , G06N3/04
摘要： Methods, systems, and apparatus, including computer programs encoded on computer storage medium, for training a neural network, wherein the neural network is configured to receive an input data item and to process the input data item to generate a respective score for each label in a predetermined set of multiple labels. The method includes actions of obtaining a set of training data that includes a plurality of training items, wherein each training item is associated with a respective label from the predetermined set of multiple labels; and modifying the training data to generate regularizing training data, comprising: for each training item, determining whether to modify the label associated with the training item, and changing the label associated with the training item to a different label from the predetermined set of labels, and training the neural network on the regularizing data.
-
公开(公告)号：US09508023B1
公开(公告)日：2016-11-29
申请号：US14257683
申请日：2014-04-21
申请人： Google Inc.
发明人： Matthew Sharifi , Sergey Ioffe , Jay Yagnik , Gheorghe Postelnicu , Dominik Roblek , George Tzanetakis
CPC分类号： G06K9/6267 , G06F17/3002 , G06F17/30244 , G06K9/00013
摘要： This disclosure relates to transformation invariant media matching. A fingerprinting component can generate a transformation invariant identifier for media content by adaptively encoding the relative ordering of interest points in media content. The interest points can be grouped into subsets, and stretch invariant descriptors can be generated for the subsets based on ratios of coordinates of interest points included in the subsets. The stretch invariant descriptors can be aggregated into a transformation invariant identifier. An identification component compares the identifier against a set of identifiers for known media content, and the media content can be matched or identified as a function of the comparison.
摘要翻译： 本公开涉及变换不变媒体匹配。 指纹分量可以通过对媒体内容中的兴趣点的相对排序进行自适应编码来生成媒体内容的变换不变标识符。 可以将兴趣点分组为子集，并且可以基于子集中包括的兴趣点坐标的比例为子集生成拉伸不变描述符。 拉伸不变描述符可以聚合成变换不变标识符。 识别部件将标识符与已知媒体内容的一组标识符进行比较，并且媒体内容可以作为比较的函数进行匹配或标识。
-
公开(公告)号：US20160216848A1
公开(公告)日：2016-07-28
申请号：US15092102
申请日：2016-04-06
申请人： Google Inc.
发明人： Sergey Ioffe , Vivek Kwatra , Matthias Grundmann
IPC分类号： G06F3/0481
CPC分类号： G06F3/0481 , G06F16/4393 , G06F16/58
摘要： A computer-implemented method, computer program product, and computing system is provided for interacting with images having similar content. In an embodiment, a method may include identifying a plurality of photographs as including a common characteristic. The method may also include generating a flipbook media item including the plurality of photographs. The method may further include associating one or more interactive control features with the flipbook media item.
摘要翻译： 提供了计算机实现的方法，计算机程序产品和计算系统，用于与具有相似内容的图像进行交互。 在一个实施例中，方法可以包括将多个照片识别为包括共同特征。 该方法还可以包括生成包括多张照片的翻页媒体项目。 该方法可以进一步包括将一个或多个交互式控制特征与翻录媒体项目相关联。
-
-
-
-
-
-
-
-
-
搜索结果
34 条记录 (耗时 0.023 秒)
