专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明申请

US20070258644A1 Pattern recognition apparatus and method therefor 审中-公开
标题翻译：模式识别装置及其方法
公开(公告)号：US20070258644A1
公开(公告)日：2007-11-08
申请号：US11712392
申请日：2007-03-01
申请人： Tomokazu Kawahara , Osamu Yamaguchi , Kenichi Maeda
发明人： Tomokazu Kawahara , Osamu Yamaguchi , Kenichi Maeda
IPC分类号： G06K9/00
CPC分类号： G06K9/00288 , G06K9/627
摘要： A pattern recognition apparatus includes an image inputting unit, a face-area extracting unit, a face-characteristic-point detecting unit, a normalized-image generating unit, a subspace generating unit, a similarity calculating unit, a reference-subspace storing unit, a judging unit, and a display unit. The pattern recognition apparatus calculates an input subspace from an input pattern, calculates a reference subspace from a reference pattern, and sets, with respect to orthogonal bases Φ1, . . . , ΦM of the input subspace and orthogonal bases Ψ1, . . . , ΨN of the reference subspace, an average of distances between Φi and Ψj (i=1, . . . , M and j=1, . . . , N) as a similarity, and performs identification using this similarity.
摘要翻译：模式识别装置包括图像输入单元，面部区域提取单元，面部特征点检测单元，归一化图像生成单元，子空间生成单元，相似度计算单元，参考子空间存储单元，判断单元和显示单元。图案识别装置从输入图案计算输入子空间，从参考图案计算参考子空间，并相对于正交基底Phi 1设定。。。，PhiM的输入子空间和正交基Psi 1，。。。，参考子空间的PsiN，Phii和Psij（i = 1，...，M和j = 1，...，N）之间的距离的平均值作为相似度，并且使用该相似性进行识别。

2. 发明授权

US08121411B2 Linear transformation matrix calculating apparatus, method thereof and program thereof 有权
标题翻译：线性变换矩阵计算装置及其方法
公开(公告)号：US08121411B2
公开(公告)日：2012-02-21
申请号：US12558817
申请日：2009-09-14
申请人： Susumu Kubota , Tomokazu Kawahara
发明人： Susumu Kubota , Tomokazu Kawahara
IPC分类号： G06K9/00 , G06F15/00 , H04N7/16
CPC分类号： G06K9/624
摘要： A linear transformation matrix calculating apparatus linearly transforms a plurality of dictionary subspaces which belong to respective categories by a linear transformation matrix respectively, selects a plurality of sets of two dictionary subspaces from the plurality of linearly transformed dictionary subspaces, calculates a loss function using similarities among the selected sets of dictionary subspaces respectively, calculates a differential parameter obtained by differentiating the loss function by the linear transformation matrix, calculates a new linear transformation matrix from the differential parameter and the linear transformation matrix by Deepest Descent Method, and updates the new linear transformation matrix as the linear transformation matrix used in the linear transformation unit.
摘要翻译：线性变换矩阵计算装置分别通过线性变换矩阵线性变换属于各个类别的多个字典子空间，从多个线性变换的字典子空间中选择多组两个字典子空间，使用相似度计算损失函数分别选择的字典子空间集合计算通过线性变换矩阵对损失函数进行微分获得的差分参数，通过深度下降法从差分参数和线性变换矩阵计算新的线性变换矩阵，并更新新的线性变换矩阵作为在线性变换单元中使用的线性变换矩阵。

3. 发明授权

US08805752B2 Learning device, learning method, and computer program product 有权
标题翻译：学习设备，学习方法和计算机程序产品
公开(公告)号：US08805752B2
公开(公告)日：2014-08-12
申请号：US13412688
申请日：2012-03-06
申请人： Tomokazu Kawahara , Tatsuo Kozakaya
发明人： Tomokazu Kawahara , Tatsuo Kozakaya
IPC分类号： G06F15/18
CPC分类号： G06K9/6269
摘要： According to an embodiment, a learning device includes a selecting unit, a learning unit, and an evaluating unit. The selecting unit performs a plurality of selection processes of selecting a plurality of groups including one or more learning samples from a learning sample storage unit, where respective learning samples are classified into any one of a plurality of categories. The learning unit learns a classification metric and obtains a set of a classification metric. The evaluating unit acquires two or more evaluation samples of different categories from an evaluation sample storage unit where respective evaluation samples are classified into any one of a plurality of categories; evaluates the classification metric included in the set of the classification metric using the two or more acquired evaluation samples; acquires a plurality of classification metric corresponding to the evaluation results from the set of the classification metric; and thereby generates an evaluation metric including the plurality of classification metric.
摘要翻译：根据实施例，学习装置包括选择单元，学习单元和评估单元。选择单元执行从学习样本存储单元中选择包括一个或多个学习样本的多个组的多个选择处理，其中各个学习样本被分类为多个类别中的任何一个。学习单元学习分类度量并获得一组分类度量。评估单元从评估样本存储单元获取不同类别的两个或更多个评估样本，其中各个评估样本被分类为多个类别中的任一个; 使用两个或多个获取的评估样本来评估包含在分类度量集合中的分类度量; 从所述分类度量的集合中获取与所述评估结果相对应的多个分类度量; 从而生成包括多个分类度量的评价度量。

4. 发明申请

US20100266166A1 IMAGE PROCESSING APPARATUS, IMAGE PROCESSING METHOD, AND STORAGE MEDIUM 失效
标题翻译：图像处理设备，图像处理方法和存储介质
公开(公告)号：US20100266166A1
公开(公告)日：2010-10-21
申请号：US12760315
申请日：2010-04-14
申请人： Tomokazu Kawahara , Tomoyuki Shibata , Tomokazu Wakasugi
发明人： Tomokazu Kawahara , Tomoyuki Shibata , Tomokazu Wakasugi
IPC分类号： G06K9/00
CPC分类号： G06K9/6807 , G06K9/00295 , G06K9/622 , G06K9/6272
摘要： An image processing apparatus includes: a sequence creating section configured to create a plurality of sequences in such a manner that one sequence includes consecutive face images of a same person in video image data; a similarity calculating section configured to calculate a first similarity of each pair in a plurality of face image dictionaries created for each sequence and a second similarity of each pair of each face image dictionary and a predetermined plurality of dictionaries; a similarity correcting section configured to correct the calculated and obtained plurality of first similarities by the second similarities; and a face clustering section configured to compare the plurality of first similarities corrected by the similarity correcting section with a predetermined threshold to cluster the plurality of face image dictionaries.
摘要翻译：一种图像处理装置，包括：序列创建部，被配置为以使得一个序列在视频图像数据中包含同一个人的连续脸部图像的方式创建多个序列; 相似度计算部，被配置为计算针对每个序列创建的多个面部图像词典中的每一对的第一相似度和每对面部图像词典和预定的多个词典的第二相似度; 相似度校正部，被配置为通过第二相似度校正所计算和获得的多个第一相似度; 以及面部聚类部，被配置为将由所述相似性校正部校正的所述多个第一相似度与预定阈值进行比较，以聚集所述多个面部图像字典。

5. 发明申请

US20100195917A1 LINER TRANSFORMATION MATRIX CALCULATING APPARATUS, METHOD THEREOF AND PROGRAM THEREOF 有权
标题翻译：线性变换矩阵计算装置及其方法及程序
公开(公告)号：US20100195917A1
公开(公告)日：2010-08-05
申请号：US12558817
申请日：2009-09-14
申请人： Susumu Kubota , Tomokazu Kawahara
发明人： Susumu Kubota , Tomokazu Kawahara
IPC分类号： G06K9/68
CPC分类号： G06K9/624
摘要： A linear transformation matrix calculating apparatus linearly transforms a plurality of dictionary subspaces which belong to respective categories by a linear transformation matrix respectively, selects a plurality of sets of two dictionary subspaces from the plurality of linearly transformed dictionary subspaces, calculates a loss function using similarities among the selected sets of dictionary subspaces respectively, calculates a differential parameter obtained by differentiating the loss function by the linear transformation matrix, calculates a new linear transformation matrix from the differential parameter and the linear transformation matrix by Deepest Descent Method, and updates the new linear transformation matrix as the linear transformation matrix used in the linear transformation unit.
摘要翻译：线性变换矩阵计算装置分别通过线性变换矩阵线性变换属于各个类别的多个字典子空间，从多个线性变换的字典子空间中选择多组两个字典子空间，使用相似度计算损失函数分别选择的字典子空间集合计算通过线性变换矩阵对损失函数进行微分获得的差分参数，通过深度下降法从差分参数和线性变换矩阵计算新的线性变换矩阵，并更新新的线性变换矩阵作为在线性变换单元中使用的线性变换矩阵。

6. 发明授权

US08428312B2 Image processing apparatus, image processing method, and storage medium 失效
标题翻译：图像处理装置，图像处理方法和存储介质
公开(公告)号：US08428312B2
公开(公告)日：2013-04-23
申请号：US12760315
申请日：2010-04-14
申请人： Tomokazu Kawahara , Tomoyuki Shibata , Tomokazu Wakusugi
发明人： Tomokazu Kawahara , Tomoyuki Shibata , Tomokazu Wakusugi
IPC分类号： G06K9/00 , G06K9/62
CPC分类号： G06K9/6807 , G06K9/00295 , G06K9/622 , G06K9/6272
摘要： An image processing apparatus includes: a sequence creating section configured to create a plurality of sequences in such a manner that one sequence includes consecutive face images of a same person in video image data; a similarity calculating section configured to calculate a first similarity of each pair in a plurality of face image dictionaries created for each sequence and a second similarity of each pair of each face image dictionary and a predetermined plurality of dictionaries; a similarity correcting section configured to correct the calculated and obtained plurality of first similarities by the second similarities; and a face clustering section configured to compare the plurality of first similarities corrected by the similarity correcting section with a predetermined threshold to cluster the plurality of face image dictionaries.
摘要翻译：一种图像处理装置，包括：序列创建部，被配置为以使得一个序列在视频图像数据中包含同一个人的连续脸部图像的方式创建多个序列; 相似度计算部，被配置为计算针对每个序列创建的多个面部图像词典中的每一对的第一相似度和每对面部图像词典和预定的多个词典的第二相似度; 相似度校正部，被配置为通过第二相似度校正所计算和获得的多个第一相似度; 以及面部聚类部，被配置为将由所述相似性校正部校正的所述多个第一相似度与预定阈值进行比较，以聚集所述多个面部图像字典。

7. 发明申请

US20120246099A1 LEARNING DEVICE, LEARNING METHOD, AND COMPUTER PROGRAM PRODUCT 有权
标题翻译：学习设备，学习方法和计算机程序产品
公开(公告)号：US20120246099A1
公开(公告)日：2012-09-27
申请号：US13412688
申请日：2012-03-06
申请人： Tomokazu Kawahara , Tatsuo Kozakaya
发明人： Tomokazu Kawahara , Tatsuo Kozakaya
IPC分类号： G06F15/18
CPC分类号： G06K9/6269
摘要： According to an embodiment, a learning device includes a selecting unit, a learning unit, and an evaluating unit. The selecting unit performs a plurality of selection processes of selecting a plurality of groups including one or more learning samples from a learning sample storage unit, where respective learning samples are classified into any one of a plurality of categories. The learning unit learns a classification metric and obtains a set of a classification metric. The evaluating unit acquires two or more evaluation samples of different categories from an evaluation sample storage unit where respective evaluation samples are classified into any one of a plurality of categories; evaluates the classification metric included in the set of the classification metric using the two or more acquired evaluation samples; acquires a plurality of classification metric corresponding to the evaluation results from the set of the classification metric; and thereby generates an evaluation metric including the plurality of classification metric.
摘要翻译：根据实施例，学习装置包括选择单元，学习单元和评估单元。选择单元执行从学习样本存储单元中选择包括一个或多个学习样本的多个组的多个选择处理，其中各个学习样本被分类为多个类别中的任何一个。学习单元学习分类度量并获得一组分类度量。评估单元从评估样本存储单元获取不同类别的两个或更多个评估样本，其中各个评估样本被分类为多个类别中的任一个; 使用两个或多个获取的评估样本来评估包含在分类度量集合中的分类度量; 从所述分类度量的集合中获取与所述评估结果相对应的多个分类度量; 从而生成包括多个分类度量的评价度量。

8. 发明授权

US08077979B2 Apparatus and method for pattern recognition 有权
标题翻译：用于模式识别的装置和方法
公开(公告)号：US08077979B2
公开(公告)日：2011-12-13
申请号：US12401300
申请日：2009-03-10
申请人： Tomokazu Kawahara , Osamu Yamaguchi
发明人： Tomokazu Kawahara , Osamu Yamaguchi
IPC分类号： G06K9/46 , H04H60/32
CPC分类号： G06K9/38 , G06K9/46
摘要： A pattern recognition method comprises steps of inputting a pattern of a recognition object performing feature extraction from the input pattern to generate a feature vector, increasing the number of quantization in an order from quantization number 1 or quantization number 2 to calculate a quantization threshold of each of the quantization number, wherein the quantization threshold of quantization number (n+1) using a quantization threshold of quantization number n (n>=1) is calculated and a quantization function having a quantization threshold corresponding to quantization number S (S>n) is generated, quantizing each component of the feature vector of the input pattern using the quantization function to generate an input quantization feature vector having each of the quantized component, storing a dictionary feature vector of the recognition object, or a quantized dictionary feature vector in which each component of the dictionary feature vector of the pattern of a recognition object is quantized; calculating a similarity between the input quantization feature vector and the dictionary feature vector, or a similarity between the input quantization feature vector and the quantized dictionary feature vector; and recognizing the recognition object based on the similarity.
摘要翻译：模式识别方法包括以下步骤：从输入模式输入执行特征提取的识别对象的图案，以生成特征向量，从量化数1或量化数2增加量化数量，以计算每个的量化阈值，其中使用量化数n（n> = 1）的量化阈值的量化数量（n + 1）的量化阈值被计算，并且具有对应于量化数S（S> n）的量化阈值的量化函数），使用量化函数量化输入图案的特征向量的每个分量，以生成具有每个量化分量，存储识别对象的词典特征向量或量化词典特征向量的输入量化特征向量字典的每个组成部分的识别对象的图案特征向量被量化计算输入量化特征向量和字典特征向量之间的相似度，或输入量化特征向量与量化词典特征向量之间的相似度; 并基于相似度识别识别对象。

9. 发明申请

US20090232399A1 APPARATUS AND METHOD FOR PATTERN RECOGNITION 有权
标题翻译：用于图案识别的装置和方法
公开(公告)号：US20090232399A1
公开(公告)日：2009-09-17
申请号：US12401300
申请日：2009-03-10
申请人： Tomokazu KAWAHARA , Osamu Yamaguchi
发明人： Tomokazu KAWAHARA , Osamu Yamaguchi
IPC分类号： G06K9/46 , G06K9/36
CPC分类号： G06K9/38 , G06K9/46
摘要： A pattern recognition method comprises steps of inputting a pattern of a recognition object performing feature extraction from the input pattern to generate a feature vector, increasing the number of quantization in an order from quantization number 1 or quantization number 2 to calculate a quantization threshold of each of the quantization number, wherein the quantization threshold of quantization number (n+1) using a quantization threshold of quantization number n (n>=1) is calculated and a quantization function having a quantization threshold corresponding to quantization number S (S>n) is generated, quantizing each component of the feature vector of the input pattern using the quantization function to generate an input quantization feature vector having each of the quantized component, storing a dictionary feature vector of the recognition object, or a quantized dictionary feature vector in which each component of the dictionary feature vector of the pattern of a recognition object is quantized; calculating a similarity between the input quantization feature vector and the dictionary feature vector, or a similarity between the input quantization feature vector and the quantized dictionary feature vector; and recognizing the recognition object based on the similarity.
摘要翻译：模式识别方法包括以下步骤：从输入模式输入执行特征提取的识别对象的图案，以生成特征向量，从量化数1或量化数2增加量化数量，以计算每个的量化阈值，其中使用量化数n（n> = 1）的量化阈值的量化数量（n + 1）的量化阈值被计算，并且具有对应于量化数S（S> n）的量化阈值的量化函数），使用量化函数量化输入图案的特征向量的每个分量，以生成具有每个量化分量，存储识别对象的词典特征向量或量化词典特征向量的输入量化特征向量字典的每个组成部分的识别对象的图案特征向量被量化计算输入量化特征向量和字典特征向量之间的相似度，或输入量化特征向量与量化词典特征向量之间的相似度; 并基于相似度识别识别对象。

10. 发明申请

US20080317350A1 PATTERN RECOGNITION APPARATUS AND METHOD 审中-公开
标题翻译：图案识别装置及方法
公开(公告)号：US20080317350A1
公开(公告)日：2008-12-25
申请号：US12139245
申请日：2008-06-13
申请人： Osamu YAMAGUCHI , Tomokazu Kawahara , Masashi Nishiyama
发明人： Osamu YAMAGUCHI , Tomokazu Kawahara , Masashi Nishiyama
IPC分类号： G06K9/46 , G06K9/00
CPC分类号： G06K9/00288 , G06K9/6247
摘要： A image recognition apparatus includes a face input unit, an input subspace calculation unit, a dictionary subspace calculation unit, an eigenvalue calculation unit, a diagonal matrix generation unit, a transformation matrix calculation unit, a transformation unit, a similarity degree calculation unit, a recognition unit. The recognition unit, based on a similarity degrees calculated by the similarity degree calculation unit, recognizes which of a plurality of categories each of a plurality of input patterns belongs to.
摘要翻译：图像识别装置包括面部输入单元，输入子空间计算单元，字典子空间计算单元，特征值计算单元，对角矩阵生成单元，变换矩阵计算单元，变换单元，相似度计算单元，识别单位。识别单元基于由相似度计算单元计算的相似度来识别多个输入模式中的每一个属于哪个类别。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式