-
公开(公告)号:US20120109858A1
公开(公告)日:2012-05-03
申请号:US12914653
申请日:2010-10-28
申请人: Ameesh Makadia , Jason E. Weston
发明人: Ameesh Makadia , Jason E. Weston
CPC分类号: G06F17/30029 , G06F17/30026 , G06F17/30047 , G06F17/3005
摘要: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for processing joint image-audio queries. In one aspect, a method includes receiving, from a client device, a joint image-audio query including query image data and query audio data. Query image feature data is determined from the query image data. Query audio feature data is determined from the audio data. The query image feature data and the query audio feature data are provided to a joint image-audio relevance model trained to generate relevance scores for a plurality of resources, each resource including resource image data defining a resource image for the resource and text data defining resource text for the resource. Each relevance score is a measure of the relevance of corresponding resource to the joint image-audio query. Data defining search results indicating the order of the resources is provided to the client device.
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,用于处理联合图像 - 音频查询。 一方面,一种方法包括从客户端设备接收包括查询图像数据和查询音频数据的联合图像 - 音频查询。 从查询图像数据确定查询图像特征数据。 从音频数据确定查询音频特征数据。 将查询图像特征数据和查询音频特征数据提供给被训练为生成多个资源的相关性分数的联合图像 - 音频相关性模型,每个资源包括定义资源的资源图像的资源图像数据和定义资源的文本数据 资源的文本。 每个相关性分数是对应资源与联合图像 - 音频查询的相关性的度量。 将指示资源顺序的搜索结果的数据提供给客户端设备。
-
公开(公告)号:US08788434B2
公开(公告)日:2014-07-22
申请号:US12914653
申请日:2010-10-28
申请人: Ameesh Makadia , Jason E. Weston
发明人: Ameesh Makadia , Jason E. Weston
IPC分类号: G06F15/18
CPC分类号: G06F17/30029 , G06F17/30026 , G06F17/30047 , G06F17/3005
摘要: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for processing joint image-audio queries. In one aspect, a method includes receiving, from a client device, a joint image-audio query including query image data and query audio data. Query image feature data is determined from the query image data. Query audio feature data is determined from the audio data. The query image feature data and the query audio feature data are provided to a joint image-audio relevance model trained to generate relevance scores for a plurality of resources, each resource including resource image data defining a resource image for the resource and text data defining resource text for the resource. Each relevance score is a measure of the relevance of corresponding resource to the joint image-audio query. Data defining search results indicating the order of the resources is provided to the client device.
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,用于处理联合图像 - 音频查询。 一方面,一种方法包括从客户端设备接收包括查询图像数据和查询音频数据的联合图像 - 音频查询。 从查询图像数据确定查询图像特征数据。 从音频数据确定查询音频特征数据。 将查询图像特征数据和查询音频特征数据提供给被训练为生成多个资源的相关性分数的联合图像 - 音频相关性模型,每个资源包括定义资源的资源图像的资源图像数据和定义资源的文本数据 资源的文本。 每个相关性分数是对应资源与联合图像 - 音频查询的相关性的度量。 将指示资源顺序的搜索结果的数据提供给客户端设备。
-
公开(公告)号:US10055493B2
公开(公告)日:2018-08-21
申请号:US13103445
申请日:2011-05-09
申请人: Geremy A. Heitz, III , Adam Berenzweig , Jason E. Weston , Ron J. Weiss , Sally A. Goldman , Thomas Walters , Samy Bengio , Douglas Eck , Jay M. Ponte , Ryan M. Rifkin
发明人: Geremy A. Heitz, III , Adam Berenzweig , Jason E. Weston , Ron J. Weiss , Sally A. Goldman , Thomas Walters , Samy Bengio , Douglas Eck , Jay M. Ponte , Ryan M. Rifkin
IPC分类号: G06F17/30
CPC分类号: G06F16/639 , G06F16/683
摘要: Generating a playlist may include designating a seed track in an audio library; identifying audio tracks in the audio library having constructs that are within a range of a corresponding construct of the seed track, where the constructs for the audio tracks are derived from frequency representations of the audio tracks, and the corresponding construct for the seed track is derived from a frequency representation of the seed track; and generating the playlist using at least some of the audio tracks that were identified.
-
公开(公告)号:US09454600B2
公开(公告)日:2016-09-27
申请号:US13363979
申请日:2012-02-01
CPC分类号: G06F17/30675 , G06F17/3028 , G06K9/52 , G06K9/6202 , G06K9/6256 , G06K9/6262 , G06K9/6296 , G06K9/66
摘要: Methods, systems and apparatus for refining image relevance models. In general, one aspect includes receiving a trained image relevance model that generates relevance measures of content feature values of images to a query, identifying a first threshold number of common content feature values for the set of training images, the common content feature values being identified as a set of content feature values that are each shared by at least a portion of the training images, identifying a subset of the set of training images having a quantity of the common content feature values greater than a second threshold number of content features values, and generating a re-trained image relevance model based on content feature values of the set of training images, wherein content feature values of the subset of training images are weighted higher than content feature values of the training images not in the subset.
摘要翻译: 图像相关模型的方法,系统和装置。 通常,一个方面包括接收经过训练的图像相关性模型,该模型生成图像的内容特征值的相关性度量到查询,识别该组训练图像的公共内容特征值的第一阈值数量,所识别的共同内容特征值 作为由训练图像的至少一部分共享的一组内容特征值,识别具有大于第二阈值数量的内容特征值的公共内容特征值的量的训练图像集合的子集, 以及基于训练图像集合的内容特征值生成重新训练的图像相关性模型,其中训练图像的子集的内容特征值被加权高于不在该子集中的训练图像的内容特征值。
-
公开(公告)号:US20120082371A1
公开(公告)日:2012-04-05
申请号:US12896318
申请日:2010-10-01
申请人: Samy Bengio , Jason E. Weston
发明人: Samy Bengio , Jason E. Weston
IPC分类号: G06K9/62
CPC分类号: G06K9/6282
摘要: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for label embedding trees for large multi-class tasks. In one aspect, a method includes mapping each image in a plurality of images and each label in a plurality of labels into a multi-dimensional label embedding space. A tree of label predictors is trained with the plurality of mapped images such that an error function is minimized in which the error function counts an error for each mapped image if any of the label predictors at any depth of the tree incorrectly predicts that the mapped image belongs to the label predictor's respective label set.
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,用于为大型多类任务标签嵌入树。 一方面,一种方法包括将多个图像中的每个图像和多个标签中的每个标签映射成多维标签嵌入空间。 使用多个映射图像训练标签预测器的树,使得误差函数被最小化,其中如果在树的任何深度处的任何标签预测器错误地预测了映射图像,则误差函数计算每个映射图像的误差 属于标签预测器的相应标签集。
-
公开(公告)号:US20150169999A1
公开(公告)日:2015-06-18
申请号:US13363979
申请日:2012-02-01
CPC分类号: G06F17/30675 , G06F17/3028 , G06K9/52 , G06K9/6202 , G06K9/6256 , G06K9/6262 , G06K9/6296 , G06K9/66
摘要: Methods, systems and apparatus for refining image relevance models. In general, one aspect includes receiving a trained image relevance model that generates relevance measures of content feature values of images to a query, identifying a first threshold number of common content feature values for the set of training images, the common content feature values being identified as a set of content feature values that are each shared by at least a portion of the training images, identifying a subset of the set of training images having a quantity of the common content feature values greater than a second threshold number of content features values, and generating a re-trained image relevance model based on content feature values of the set of training images, wherein content feature values of the subset of training images are weighted higher than content feature values of the training images not in the subset.
摘要翻译: 图像相关模型的方法,系统和装置。 通常,一个方面包括接收经过训练的图像相关性模型,该模型生成图像的内容特征值的相关性度量到查询,识别该组训练图像的公共内容特征值的第一阈值数量,所识别的共同内容特征值 作为由训练图像的至少一部分共享的一组内容特征值,识别具有大于第二阈值数量的内容特征值的公共内容特征值的量的训练图像集合的子集, 以及基于训练图像集合的内容特征值生成重新训练的图像相关性模型,其中训练图像的子集的内容特征值被加权高于不在该子集中的训练图像的内容特征值。
-
公开(公告)号:US20120290621A1
公开(公告)日:2012-11-15
申请号:US13103445
申请日:2011-05-09
申请人: Geremy A. Heitz, III , Adam Berenzweig , Jason E. Weston , Ron J. Weiss , Sally A. Goldman , Thomas Walters , Samy Bengio , Douglas Eck , Jay M. Ponte , Ryan M. Rifkin
发明人: Geremy A. Heitz, III , Adam Berenzweig , Jason E. Weston , Ron J. Weiss , Sally A. Goldman , Thomas Walters , Samy Bengio , Douglas Eck , Jay M. Ponte , Ryan M. Rifkin
IPC分类号: G06F17/30
CPC分类号: G06F17/30772 , G06F17/30743
摘要: Generating a playlist may include designating a seed track in an audio library; identifying audio tracks in the audio library having constructs that are within a range of a corresponding construct of the seed track, where the constructs for the audio tracks are derived from frequency representations of the audio tracks, and the corresponding construct for the seed track is derived from a frequency representation of the seed track; and generating the playlist using at least some of the audio tracks that were identified.
摘要翻译: 生成播放列表可以包括在音频库中指定种子轨道; 识别具有在种子轨道的相应构造的范围内的构造的音频库中的音轨,其中音频轨道的构造从音频轨道的频率表示导出,并且导出用于种子轨道的相应构造 从种子轨迹的频率表示; 以及使用所识别的至少一些音轨生成播放列表。
-
-
-
-
-
-