-
公开(公告)号：US08566270B2
公开(公告)日：2013-10-22
申请号：US13242145
申请日：2011-09-23
CPC分类号： G06F17/2715 , G06F17/30707
摘要： A sparse representation method of text classification is described. An input text document is represented as a document feature vector y. A category dictionary H provides possible examples [h1; h2; . . . ; hn] of the document feature vector y. The input text document is classified using a sparse representation text classification algorithm that solves for y=Hβ where a sparseness condition is enforced on β to select a small number of examples from the dictionary H to describe the document feature vector y.
摘要翻译： 描述了文本分类的稀疏表示方法。 输入文本文档表示为文档特征向量y。 类别字典H提供了可能的例子[h1; h2; 。 。 。 ; hn]的文档特征向量y。 输入文本文档使用稀疏表示文本分类算法进行分类，该算法解决了y = Hbeta，其中在beta上执行稀疏条件以从字典H中选择少量示例来描述文档特征向量y。
-
公开(公告)号：US20120078834A1
公开(公告)日：2012-03-29
申请号：US13242145
申请日：2011-09-23
IPC分类号： G06N5/02
CPC分类号： G06F17/2715 , G06F17/30707
摘要： A sparse representation method of text classification is described. An input text document is represented as a document feature vector y. A category dictionary H provides possible examples [h1; h2; . . . ; hn] of the document feature vector y. The input text document is classified using a sparse representation text classification algorithm that solves for y=Hβ where a sparseness condition is enforced on β to select a small number of examples from the dictionary H to describe the document feature vector y.
摘要翻译： 描述了文本分类的稀疏表示方法。 输入文本文档表示为文档特征向量y。 类别字典H提供了可能的例子[h1; h2; 。 。 。 ; hn]的文档特征向量y。 输入文本文档使用解决y = H＆bgr的稀疏表示文本分类算法进行分类; 在“bgr”上执行稀疏条件 从字典H中选择少量示例来描述文档特征向量y。
-
公开(公告)号：US20140019857A1
公开(公告)日：2014-01-16
申请号：US13565360
申请日：2012-08-02
IPC分类号： G06F17/24
摘要： An embodiment of the invention directed to a method is associated with data content, comprising discrete data portions including first data and second data portions separated from each other in the data content. A copy operation is implemented on data portions so that at least some of the data portions are each copied to a buffer, which include the first and second data portions. A paste operation is carried out to present each of the copied data portions as an input for an output data selection task. Prespecified criteria is used in the output data selection task to select a number of the copied data portions to be selected data for a given purpose, the selected number of copied data portions being less than data portions presented by the paste operation, and the selected copied data portions including the first and second data portions.
摘要翻译： 涉及一种方法的本发明的实施例与数据内容相关联，包括离散数据部分，其包括在数据内容中彼此分离的第一数据和第二数据部分。 在数据部分上实现复制操作，使得至少一些数据部分被复制到包括第一和第二数据部分的缓冲器中。 执行粘贴操作以将每个复制数据部分呈现为输出数据选择任务的输入。 在输出数据选择任务中使用预定标准来选择用于给定目的的要被选择数据的复制数据部分的数量，所选择的复制数据部分数量小于通过粘贴操作呈现的数据部分，并且所选择的复制 数据部分包括第一和第二数据部分。
-
公开(公告)号：US08484024B2
公开(公告)日：2013-07-09
申请号：US13034293
申请日：2011-02-24
IPC分类号： G10L15/06
CPC分类号： G10L15/063 , G10L15/02 , G10L15/083 , G10L2015/025 , G10L2015/0631
摘要： Techniques are disclosed for using phonetic features for speech recognition. For example, a method comprises the steps of obtaining a first dictionary and a training data set associated with a speech recognition system, computing one or more support parameters from the training data set, transforming the first dictionary into a second dictionary, wherein the second dictionary is a function of one or more phonetic labels of the first dictionary, and using the one or more support parameters to select one or more samples from the second dictionary to create a set of one or more exemplar-based class identification features for a pattern recognition task.
摘要翻译： 公开了使用语音特征进行语音识别的技术。 例如，一种方法包括以下步骤：获得与语音识别系统相关联的第一字典和训练数据集，从训练数据集计算一个或多个支持参数，将第一字典变换为第二字典，其中第二字典 是第一字典的一个或多个语音标签的功能，并且使用一个或多个支持参数从第二字典中选择一个或多个样本，以创建用于模式识别的一个或多个基于样本的类识别特征的集合 任务。
-
公开(公告)号：US20120221333A1
公开(公告)日：2012-08-30
申请号：US13034293
申请日：2011-02-24
IPC分类号： G10L15/06
CPC分类号： G10L15/063 , G10L15/02 , G10L15/083 , G10L2015/025 , G10L2015/0631
摘要： Techniques are disclosed for using phonetic features for speech recognition. For example, a method comprises the steps of obtaining a first dictionary and a training data set associated with a speech recognition system, computing one or more support parameters from the training data set, transforming the first dictionary into a second dictionary, wherein the second dictionary is a function of one or more phonetic labels of the first dictionary, and using the one or more support parameters to select one or more samples from the second dictionary to create a set of one or more exemplar-based class identification features for a pattern recognition task.
摘要翻译： 公开了使用语音特征进行语音识别的技术。 例如，一种方法包括以下步骤：获得与语音识别系统相关联的第一字典和训练数据集，从训练数据集计算一个或多个支持参数，将第一字典变换为第二字典，其中第二字典 是第一字典的一个或多个语音标签的功能，并且使用一个或多个支持参数从第二字典中选择一个或多个样本，以创建用于模式识别的一个或多个基于样本的类识别特征的集合 任务。
-
公开(公告)号：US20110282925A1
公开(公告)日：2011-11-17
申请号：US12777768
申请日：2010-05-11
CPC分类号： G06F17/11
摘要： An optimization system and method includes determining a best gradient as a sparse direction in a function having a plurality of parameters. The sparse direction includes a direction that maximizes change of the function. This maximum change of the function is determined by performing an optimization process that gives maximum growth subject to a sparsity regularized constraint. An extended Baum Welch (EBW) method can be used to identify the sparse direction. A best step size is determined along the sparse direction by finding magnitudes of entries of direction that maximizes the function restricted to the sparse direction. A solution is recursively refined for the function optimization using a processor and storage media.
摘要翻译： 优化系统和方法包括在具有多个参数的函数中确定最佳梯度作为稀疏方向。 稀疏方向包括使功能变化最大化的方向。 通过执行优化处理来确定功能的最大变化，该优化过程允许受到稀疏正则化约束的最大增长。 扩展的Baum Welch（EBW）方法可用于识别稀疏方向。 通过找到使限于稀疏方向的功能最大化的方向条目的大小，沿着稀疏方向确定最佳步长。 使用处理器和存储介质递归地优化了功能优化的解决方案。
-
公开(公告)号：US08856004B2
公开(公告)日：2014-10-07
申请号：US13107001
申请日：2011-05-13
申请人： Sara H. Basson , Rick Hamilton , Dan Ning Jiang , Dimitri Kanevsky , David Nahamoo , Michael Picheny , Bhuvana Ramabhadran , Tara N. Sainath
发明人： Sara H. Basson , Rick Hamilton , Dan Ning Jiang , Dimitri Kanevsky , David Nahamoo , Michael Picheny , Bhuvana Ramabhadran , Tara N. Sainath
IPC分类号： G10L15/26
CPC分类号： G06F17/28 , G06F17/2785 , G06F17/289 , G10L15/063 , G10L15/26
摘要： Techniques for converting spoken speech into written speech are provided. The techniques include transcribing input speech via speech recognition, mapping each spoken utterance from input speech into a corresponding formal utterance, and mapping each formal utterance into a stylistically formatted written utterance.
-
公开(公告)号：US08527566B2
公开(公告)日：2013-09-03
申请号：US12777768
申请日：2010-05-11
IPC分类号： G06F7/00
CPC分类号： G06F17/11
摘要： An optimization system and method includes determining a best gradient as a sparse direction in a function having a plurality of parameters. The sparse direction includes a direction that maximizes change of the function. This maximum change of the function is determined by performing an optimization process that gives maximum growth subject to a sparsity regularized constraint. An extended Baum Welch (EBW) method can be used to identify the sparse direction. A best step size is determined along the sparse direction by finding magnitudes of entries of direction that maximizes the function restricted to the sparse direction. A solution is recursively refined for the function optimization using a processor and storage media.
摘要翻译： 优化系统和方法包括在具有多个参数的函数中确定最佳梯度作为稀疏方向。 稀疏方向包括使功能变化最大化的方向。 通过执行优化处理来确定功能的最大变化，该优化过程允许受到稀疏正则化约束的最大增长。 扩展的Baum Welch（EBW）方法可用于识别稀疏方向。 通过找到使限于稀疏方向的功能最大化的方向条目的大小，沿着稀疏方向确定最佳步长。 使用处理器和存储介质递归地优化了功能优化的解决方案。
-
公开(公告)号：US20120078621A1
公开(公告)日：2012-03-29
申请号：US12889845
申请日：2010-09-24
IPC分类号： G10L15/00
CPC分类号： G10L15/02
摘要： Techniques are disclosed for generating and using sparse representation features to improve speech recognition performance. In particular, principles of the invention provide sparse representation exemplar-based recognition techniques. For example, a method comprises the following steps. A test vector and a training data set associated with a speech recognition system are obtained. A subset of the training data set is selected. The test vector is mapped with the selected subset of the training data set as a linear combination that is weighted by a sparseness constraint such that a new test feature set is formed wherein the training data set is moved more closely to the test vector subject to the sparseness constraint. An acoustic model is trained on the new test feature set.The acoustic model trained on the new test feature set may be used to decode user speech input to the speech recognition system.
摘要翻译： 公开了用于生成和使用稀疏表示特征以改善语音识别性能的技术。 特别地，本发明的原理提供了基于示例的稀疏表示识别技术。 例如，一种方法包括以下步骤。 获得与语音识别系统相关联的测试向量和训练数据集。 选择训练数据集的子集。 将测试向量与所选择的训练数据集的子集映射为由稀疏约束加权的线性组合，使得形成新的测试特征集合，其中训练数据集更接近地移动到受测对象的测试向量 稀疏约束 在新的测试功能集上训练声学模型。 在新测试特征集上训练的声学模型可以用于解码输入到语音识别系统的用户语音。
-
公开(公告)号：US08484023B2
公开(公告)日：2013-07-09
申请号：US12889845
申请日：2010-09-24
IPC分类号： G10L15/06
CPC分类号： G10L15/02
摘要： Techniques are disclosed for generating and using sparse representation features to improve speech recognition performance. In particular, principles of the invention provide sparse representation exemplar-based recognition techniques. For example, a method comprises the following steps. A test vector and a training data set associated with a speech recognition system are obtained. A subset of the training data set is selected. The test vector is mapped with the selected subset of the training data set as a linear combination that is weighted by a sparseness constraint such that a new test feature set is formed wherein the training data set is moved more closely to the test vector subject to the sparseness constraint. An acoustic model is trained on the new test feature set. The acoustic model trained on the new test feature set may be used to decode user speech input to the speech recognition system.
摘要翻译： 公开了用于生成和使用稀疏表示特征以改善语音识别性能的技术。 特别地，本发明的原理提供了基于示例的稀疏表示识别技术。 例如，一种方法包括以下步骤。 获得与语音识别系统相关联的测试向量和训练数据集。 选择训练数据集的子集。 将测试向量与所选择的训练数据集的子集映射为由稀疏约束加权的线性组合，使得形成新的测试特征集合，其中训练数据集更接近地移动到受测对象的测试向量 稀疏约束 在新的测试功能集上训练声学模型。 在新测试特征集上训练的声学模型可以用于解码输入到语音识别系统的用户语音。
-
-
-
-
-
-
-
-
-
搜索结果
51 条记录 (耗时 0.023 秒)
