公开(公告)号：US20110222785A1

公开(公告)日：2011-09-15

申请号：US12923277

申请日：2010-09-13

申请人： Makoto Hirohata ,  Tomoyuki Shibata

发明人： Makoto Hirohata ,  Tomoyuki Shibata

IPC分类号： G06K9/62

CPC分类号： G06K9/6226 ,  G06K9/00288 ,  G06K9/6276

摘要： An acquisition unit acquires feature vectors (the number is N). A first selection unit selects first neighbor features (k (1≦k≦N)) of each feature vector in order of higher similarity from the feature vectors. A second selection unit generates a plurality of groups each including similar feature vectors from the feature vectors, and selects second neighbor features (u (1≦k+u≦N−2)) of each feature vector in order of higher similarity. Each of the second neighbor features is differently included in a group. A determination unit calculates a density of each feature vector by using a threshold, the first neighbor features and the second neighbor features, and determines feature vectors to be classified into the same class as each feature vector by using the density and a threshold. A classification unit classifies the feature vectors into a plurality of classes by using the selection result. A control unit controls each threshold.

摘要翻译： 采集单元获取特征向量（数量为N）。 第一选择单元按照与特征向量相似度更高的顺序选择每个特征向量的第一邻近特征（k（1＆nlE; k＆nlE; N））。 第二选择单元生成包括来自特征向量的类似特征向量的多个组，并且以相似度较高的顺序选择每个特征向量的第二邻居特征（u（1＆nlE; k + u＆nlE; N-2））。 每个第二邻居特征被不同地包括在一个组中。 确定单元通过使用阈值，第一相邻特征和第二相邻特征来计算每个特征向量的密度，并且通过使用密度和阈值来确定要分类为与每个特征向量相同类别的特征向量。 分类单元通过使用选择结果将特征向量分类为多个类。 控制单元控制每个阈值。

公开(公告)号：US09384563B2

公开(公告)日：2016-07-05

申请号：US13535952

申请日：2012-06-28

申请人： Shihomi Takahashi ,  Masashi Nishiyama ,  Tomoyuki Shibata ,  Makoto Hirohata ,  Masahiro Sekine ,  Kaoru Sugita

发明人： Shihomi Takahashi ,  Masashi Nishiyama ,  Tomoyuki Shibata ,  Makoto Hirohata ,  Masahiro Sekine ,  Kaoru Sugita

IPC分类号： H04N7/18 ,  G06T11/00 ,  G06K9/50 ,  G06K9/48

CPC分类号： G06T11/00 ,  G06K9/48 ,  G06K9/50

摘要： In an embodiment, an image processing device includes: a first calculating unit; a selecting unit; an extracting unit; a second calculating unit; a determining unit; and a generating unit. The first calculating unit calculates posture information of a subject from a first image including the subject. The selecting unit selects a second image associated with the posture information from a plurality of second images including clothing. The extracting unit extracts a first feature of the subject from a first contour in the first image and extracts a second feature from a second contour in the second image. The second calculating unit calculates a first difference between the first feature and the second feature. The determining unit determines a transparency that is higher as the first difference is larger. The generating unit generates a composite image combining the first image and the second image with the determined transparency.

摘要翻译： 在一个实施例中，图像处理装置包括：第一计算单元; 选择单元; 提取单元; 第二计算单元; 决定单位 和发电机组。 第一计算单元从包括对象的第一图像计算被摄体的姿势信息。 选择单元从包括衣服的多个第二图像中选择与姿势信息相关联的第二图像。 提取单元从第一图像中的第一轮廓提取对象的第一特征并从第二图像中的第二轮廓提取第二特征。 第二计算单元计算第一特征和第二特征之间的第一差值。 确定单元确定随着第一差异较大而更高的透明度。 生成单元生成以确定的透明度组合第一图像和第二图像的合成图像。

公开(公告)号：US08675973B2

公开(公告)日：2014-03-18

申请号：US12923277

申请日：2010-09-13

申请人： Makoto Hirohata ,  Tomoyuki Shibata

发明人： Makoto Hirohata ,  Tomoyuki Shibata

IPC分类号： G06K9/62

CPC分类号： G06K9/6226 ,  G06K9/00288 ,  G06K9/6276

摘要： An acquisition unit acquires feature vectors (the number is N). A first selection unit selects first neighbor features (k (1≦k≦N)) of each feature vector in order of higher similarity from the feature vectors. A second selection unit generates a plurality of groups each including similar feature vectors from the feature vectors, and selects second neighbor features (u (1≦k+u≦N−2)) of each feature vector in order of higher similarity. Each of the second neighbor features is differently included in a group. A determination unit calculates a density of each feature vector by using a threshold, the first neighbor features and the second neighbor features, and determines feature vectors to be classified into the same class as each feature vector by using the density and a threshold. A classification unit classifies the feature vectors into a plurality of classes by using the selection result. A control unit controls each threshold.

摘要翻译： 采集单元获取特征向量（数量为N）。 第一选择单元按照与特征向量相似度较高的顺序选择每个特征向量的第一邻近特征（k（1 @ k @ N））。 第二选择单元生成包括来自特征向量的类似特征向量的多个组，并且以较高相似度的顺序选择每个特征向量的第二邻居特征（u（1 @ k + u @ N-2））。 每个第二邻居特征被不同地包括在一个组中。 确定单元通过使用阈值，第一相邻特征和第二相邻特征来计算每个特征向量的密度，并且通过使用密度和阈值来确定要分类为与每个特征向量相同类别的特征向量。 分类单元通过使用选择结果将特征向量分类为多个类。 控制单元控制每个阈值。

公开(公告)号：US08145486B2

公开(公告)日：2012-03-27

申请号：US12007379

申请日：2008-01-09

申请人： Makoto Hirohata

发明人： Makoto Hirohata

IPC分类号： G10L17/00

CPC分类号： G10L17/00 ,  G10L15/07

摘要： Acoustic models to provide features to a speech signal are created based on speech features included in regions where similarities of acoustic models created based on speech features in a certain time length are equal to or greater than a predetermined value. Feature vectors acquired by using the acoustic models of the regions and the speech features to provide features to speech signals of second segments are grouped by speaker.

摘要翻译： 基于语音特征产生声学模型，该语音特征包括在基于特定时间长度的语音特征所产生的声学模型的相似性等于或大于预定值的区域中。 通过使用区域的声学模型和语音特征获得的向第二段的语音信号提供特征的特征矢量由扬声器分组。

公开(公告)号：US20100079673A1

公开(公告)日：2010-04-01

申请号：US12382567

申请日：2009-03-18

申请人： Koji Yamamoto ,  Shunsuke Takayama ,  Hisashi Aoki ,  Yoshihiro Ohmori ,  Kazunori Imoto ,  Shuta Ogawa ,  Makoto Hirohata ,  Shigeru Motoi

发明人： Koji Yamamoto ,  Shunsuke Takayama ,  Hisashi Aoki ,  Yoshihiro Ohmori ,  Kazunori Imoto ,  Shuta Ogawa ,  Makoto Hirohata ,  Shigeru Motoi

IPC分类号： H04N5/14

CPC分类号： H04N5/147

摘要： A video processing apparatus according to the invention detects telops displayed in an entered video, selects specific telops which satisfy arbitrary conditions from among the telops, acquires a plurality of specific telops within an arbitrary time range as one group from among the plurality of specific telops, coordinates two of the specific telops from the group, and extracts a specific segment interposed between the two of the specific telops.

摘要翻译： 根据本发明的视频处理装置检测在输入的视频中显示的字幕，从字幕中选择满足任意条件的特定字幕，从多个特定字幕中获取任意时间范围内的多个特定字幕作为一组， 协调来自组中的两个特定字幕，并提取插入在两个特定字幕之间的特定段。

公开(公告)号：US08879788B2

公开(公告)日：2014-11-04

申请号：US13240278

申请日：2011-09-22

申请人： Koji Yamamoto ,  Makoto Hirohata

发明人： Koji Yamamoto ,  Makoto Hirohata

IPC分类号： G06K9/00 ,  H04N21/233 ,  H04N21/8549 ,  H04N21/234 ,  H04N21/25 ,  H04N21/258 ,  H04N21/845

CPC分类号： H04N21/233 ,  G06K9/00711 ,  H04N21/23418 ,  H04N21/252 ,  H04N21/25891 ,  H04N21/8456 ,  H04N21/8549

摘要： According to one embodiment, a video processing apparatus includes an acquisition unit, a first extraction unit, a generation unit, a second extraction unit, a computation unit and a selection unit. The acquisition unit is configured to acquire video streams. A first extraction unit is configured to analyze at least one of the moving pictures and the sounds for each video stream and to extract feature values. A generation unit is configured to generate segments by dividing each video stream, and to generate associated segment groups. A second extraction unit is configured to extract the associated segment groups that number of associated segments is greater than or equal to threshold as common video segment groups. A computation unit is configured to compute summarization score. A selection unit is configured to select segments used for a summarized video as summarization segments from the common video segment groups based on the summarization score.

摘要翻译： 根据一个实施例，视频处理装置包括获取单元，第一提取单元，生成单元，第二提取单元，计算单元和选择单元。 采集单元被配置为获取视频流。 第一提取单元被配置为分析每个视频流的运动图像和声音中的至少一个并提取特征值。 生成单元被配置为通过划分每个视频流来生成段，并且生成相关联的段组。 第二提取单元被配置为提取相关联的片段组，所述相关联的段的数量大于或等于阈值作为公共视频片段组。 计算单元被配置为计算汇总分数。 选择单元被配置为基于总结分数从公共视频段组中选择用于汇总视频的段作为汇总段。

公开(公告)号：US08713030B2

公开(公告)日：2014-04-29

申请号：US13376274

申请日：2009-06-05

申请人： Kazunori Imoto ,  Makoto Hirohata ,  Hisashi Aoki

发明人： Kazunori Imoto ,  Makoto Hirohata ,  Hisashi Aoki

IPC分类号： G06F17/30

CPC分类号： G06F17/30787 ,  G11B27/034 ,  G11B27/28 ,  H04N5/76

摘要： A video editing apparatus 100 includes a registering unit 91 configured to register a key candidate having a feature vector of a sound signal which is determined to be registered on the basis of a co-occurrence score to a managing unit 51 as a search key, and a cutting out unit 71 configured to obtain an integration score in each of the blocks from the degree of similarity of the registered search key in each of the blocks and cut out a group of blocks exceeding an integration threshold value from among the integration scores as one video scene.

摘要翻译： 视频编辑装置100包括：注册单元91，被配置为将具有基于共有记录被确定登记的声音信号的特征向量的关键候选者登记到作为搜索关键字的管理单元51;以及 切出单元71，被配置为从每个块中的登记的搜索关键字的相似度获得每个块中的积分分数，并从积分分数中除去超过积分阈值的一组块，作为一个 视频场景。

公开(公告)号：US20120237042A1

公开(公告)日：2012-09-20

申请号：US13423631

申请日：2012-03-19

申请人： Makoto Hirohata ,  Kazunori Imoto ,  Hisashi Aoki

发明人： Makoto Hirohata ,  Kazunori Imoto ,  Hisashi Aoki

IPC分类号： H04R29/00

CPC分类号： G10L25/78

摘要： In an example signal clustering apparatus, a feature of a signal is divided into segments. A first feature vector of each segment is calculated, the first feature vector having has a plurality of elements corresponding to each reference model. A value of an element attenuates when a feature of the segment shifts from a center of a distribution of the reference model corresponding to the element. A similarity between two reference models is calculated. A second feature vector of each segment is calculated, the second feature vector having a plurality of elements corresponding to each reference model. A value of an element is a weighted sum and segments of second feature vectors of which the plurality of elements are similar values are clustered to one class.

摘要翻译： 在示例信号聚类设备中，信号的特征被分成多个段。 计算每个段的第一特征向量，第一特征向量具有对应于每个参考模型的多个元素。 当元素的特征从对应于元素的参考模型的分布的中心偏移时，元素的值衰减。 计算两个参考模型之间的相似度。 计算每个段的第二特征向量，第二特征向量具有对应于每个参考模型的多个元素。 元素的值是加权和，并且多个元素是相似值的第二特征向量的集合到一个类。

公开(公告)号：US20120117087A1

公开(公告)日：2012-05-10

申请号：US13376274

申请日：2009-06-05

申请人： Kazunori Imoto ,  Makoto Hirohata ,  Hisashi Aoki

发明人： Kazunori Imoto ,  Makoto Hirohata ,  Hisashi Aoki

IPC分类号： G06F17/30

CPC分类号： G06F17/30787 ,  G11B27/034 ,  G11B27/28 ,  H04N5/76

摘要： A video editing apparatus 100 includes a registering unit 91 configured to register a key candidate having a feature vector of a sound signal which is determined to be registered on the basis of a co-occurrence score to a managing unit 51 as a search key, and a cutting out unit 71 configured to obtain an integration score in each of the blocks from the degree of similarity of the registered search key in each of the blocks and cut out a group of blocks exceeding an integration threshold value from among the integration scores as one video scene.

摘要翻译： 视频编辑装置100包括：注册单元91，被配置为将具有基于共有记录被确定登记的声音信号的特征向量的关键候选者登记到作为搜索关键字的管理单元51;以及 切出单元71，被配置为从每个块中的登记的搜索关键字的相似度获得每个块中的积分分数，并从积分分数中除去超过积分阈值的一组块，作为一个 视频场景。

公开(公告)号：US08804973B2

公开(公告)日：2014-08-12

申请号：US13423631

申请日：2012-03-19

申请人： Makoto Hirohata ,  Kazunori Imoto ,  Hisashi Aoki

发明人： Makoto Hirohata ,  Kazunori Imoto ,  Hisashi Aoki

IPC分类号： H04R29/00

CPC分类号： G10L25/78

摘要： In an example signal clustering apparatus, a feature of a signal is divided into segments. A first feature vector of each segment is calculated, the first feature vector having has a plurality of elements corresponding to each reference model. A value of an element attenuates when a feature of the segment shifts from a center of a distribution of the reference model corresponding to the element. A similarity between two reference models is calculated. A second feature vector of each segment is calculated, the second feature vector having a plurality of elements corresponding to each reference model. A value of an element is a weighted sum and segments of second feature vectors of which the plurality of elements are similar values are clustered to one class.

摘要翻译： 在示例信号聚类设备中，信号的特征被分成多个段。 计算每个段的第一特征向量，第一特征向量具有对应于每个参考模型的多个元素。 当元素的特征从对应于元素的参考模型的分布的中心偏移时，元素的值衰减。 计算两个参考模型之间的相似度。 计算每个段的第二特征向量，第二特征向量具有对应于每个参考模型的多个元素。 元素的值是加权和，并且多个元素是相似值的第二特征向量的集合到一个类。