公开(公告)号：US20120033853A1

公开(公告)日：2012-02-09

申请号：US13194640

申请日：2011-07-29

申请人： Yuji Kaneda ,  Masakazu Matsugu ,  Katsuhiko Mori

发明人： Yuji Kaneda ,  Masakazu Matsugu ,  Katsuhiko Mori

IPC分类号： G06K9/46

CPC分类号： G06K9/00597 ,  G06K9/00221

摘要： The present invention refers to an information processing apparatus comprising: an obtaining unit adapted to obtain an image of an object; a face region detection unit adapted to detect a face region of the object from the image; an eye region detection unit adapted to detect an eye region of the object; a generation unit adapted to generate a high-resolution image and low-resolution image of the face region detected by the face region detection means; a first extraction unit adapted to extract a first feature amount indicating a direction of a face existing in the face region from the low-resolution image; a second extraction unit adapted to extract a second feature amount indicating a direction of an eye existing in the eye region from the high-resolution image; and an estimation unit adapted to estimate a gaze direction of the object from the first feature amount and the second feature amount.

摘要翻译： 本发明涉及一种信息处理装置，包括：获取单元，用于获得对象的图像; 面部区域检测单元，适于从图像检测对象的面部区域; 眼部区域检测单元，其适于检测所述物体的眼睛区域; 生成单元，其适于生成由所述面部区域检测单元检测出的面部区域的高分辨率图像和低分辨率图像; 第一提取单元，适于从低分辨率图像中提取指示存在于面部区域中的脸部的方向的第一特征量; 第二提取单元，适于从高分辨率图像提取指示存在于眼睛区域中的眼睛的方向的第二特征量; 以及估计单元，其适于从所述第一特征量和所述第二特征量估计所述对象的凝视方向。

公开(公告)号：US20110158536A1

公开(公告)日：2011-06-30

申请号：US12973172

申请日：2010-12-20

申请人： Shunsuke Nakano ,  Hiroshi Sato ,  Katsuhiko Mori

发明人： Shunsuke Nakano ,  Hiroshi Sato ,  Katsuhiko Mori

IPC分类号： G06K9/46

CPC分类号： G06K9/00281 ,  G06K9/00926

摘要： An object identification apparatus, which identifies a class of a plurality of registered classes to which an object included in an input image belongs, registers a plurality of registered images used in identification of an input image, and holds designation data required to designate partial areas used in identification. The apparatus extracts feature amounts of the partial areas designated by the designation data from the input image and each of the plurality of registered images, and identifies a class to which the input image belongs based on the extracted feature amounts. The apparatus determines based on the plurality of registered images whether or not is required to update processing contents of an identification unit. When it is determined that the processing contents are required to be updated, the apparatus updates the designation data based on registered images.

摘要翻译： 一种对象识别装置，其识别包含在输入图像中的对象所属的多个注册类别的类别，登记在输入图像的识别中使用的多个登记图像，并保存指定使用的部分区域所需的指定数据 在识别。 该装置从输入图像和多个登记图像中的每一个提取由指定数据指定的部分区域的特征量，并且基于提取的特征量来识别输入图像所属的类别。 该装置基于多个登记图像确定是否需要更新识别单元的处理内容。 当确定需要更新处理内容时，装置基于登记图像来更新指定数据。

公开(公告)号：US07864228B2

公开(公告)日：2011-01-04

申请号：US11258694

申请日：2005-10-25

申请人： Katsuhiko Mori ,  Masakazu Matsugu ,  Yusuke Mitarai

发明人： Katsuhiko Mori ,  Masakazu Matsugu ,  Yusuke Mitarai

IPC分类号： H04N5/262 ,  H04N5/222

CPC分类号： H04N5/23293

摘要： An object area in an image is determined, and an area image representing the object area is superimposed on the object in the image. In this image, an area to be photographed with high quality is specified using the area image, and a control parameter for controlling an image pickup unit is adjusted so that the quality level of an image within the area to be photographed with high quality is increased to a desired level. Thus, when photographing an image, an area to be photographed with high quality can be easily set in the image.

摘要翻译： 确定图像中的对象区域，并且将表示对象区域的区域图像叠加在图像中的对象上。 在该图像中，使用区域图像指定要拍摄的高品质区域，并且调整用于控制图像拾取单元的控制参数，使得高质量的要拍摄的区域内的图像的质量水平增加 达到所需的水平。 因此，当拍摄图像时，可以容易地在图像中设置要拍摄的高质量区域。

公开(公告)号：US07577297B2

公开(公告)日：2009-08-18

申请号：US10539882

申请日：2003-12-16

申请人： Katsuhiko Mori ,  Masakazu Matsugu ,  Mie Ishii ,  Yusuke Mitarai

发明人： Katsuhiko Mori ,  Masakazu Matsugu ,  Mie Ishii ,  Yusuke Mitarai

IPC分类号： G06K9/46

CPC分类号： G06K9/00

摘要： In a pattern identification device performing pattern identification of data which has been input by hierarchically extracting features, a primary feature is extracted and distribution of at least one feature extraction result extracted is analyzed, a secondary feature is extracted according to this analysis result. Thus, it is possible to perform pattern identification robustly against fluctuations of the input pattern at a reduced processing cost while reducing the probability of occurrence of incorrect identification.

摘要翻译： 在通过分层抽取特征输入的数据的图案识别装置的图案识别装置中，提取主要特征并分析提取的至少一个特征提取结果的分布，根据该分析结果提取次要特征。 因此，可以以降低的处理成本对输入图案的波动进行鲁棒性的图案识别，同时降低错误识别的发生概率。

公开(公告)号：US20090003709A1

公开(公告)日：2009-01-01

申请号：US12147123

申请日：2008-06-26

申请人： Yuji Kaneda ,  Katsuhiko Mori ,  Hiroshi Sato

发明人： Yuji Kaneda ,  Katsuhiko Mori ,  Hiroshi Sato

IPC分类号： G06K9/46

CPC分类号： G06K9/00845 ,  G06K9/00315 ,  G06K9/0061 ,  G06K9/4628

摘要： An image-processing apparatus for executing accurate facial expression recognition even for a subject hard to recognize a facial expression is provided. A person's face region is extracted from an image input from an image input unit. A predetermined partial region that changes between when the facial expression is in the first and second states is extracted from the extracted face region. A facial expression evaluation value is calculated using an evaluation value calculation formula. When the calculated facial expression evaluation value exceeds a threshold value, it is determined that the facial expression is in the second state. If the difference between the maximum value and the minimum value of the calculated facial expression evaluation value within a predetermined time is smaller than a predetermined value, the evaluation value calculation formula or its parameter is changed to increase the difference.

摘要翻译： 提供了即使对于难以识别面部表情的对象也执行准确的面部表情识别的图像处理装置。 从图像输入单元输入的图像中提取人脸区域。 从所提取的面部区域中提取在面部表情处于第一状态和第二状态之间时的变化的预定部分区域。 使用评价值计算公式计算面部表情评价值。 当所计算的面部表情评估值超过阈值时，确定面部表情处于第二状态。 如果计算出的面部表情评估值的最大值与最小值之间的差在预定时间内小于预定值，则改变评估值计算公式或其参数以增加差异。

公开(公告)号：US20080212098A1

公开(公告)日：2008-09-04

申请号：US11961982

申请日：2007-12-20

申请人： Hiroshi Sato ,  Masakazu Matsugu ,  Katsuhiko Mori ,  Kan Torii ,  Yusuke Mitarai

发明人： Hiroshi Sato ,  Masakazu Matsugu ,  Katsuhiko Mori ,  Kan Torii ,  Yusuke Mitarai

IPC分类号： G01B11/00

CPC分类号： G06K9/00228 ,  G06K9/3241

摘要： A method includes detecting a feature of an input pattern using a plurality of feature detectors, selecting at least one of the feature detectors based on their output values, and calculating a feature quantity of the input pattern based on an output value from at least one selected feature detector.

摘要翻译： 一种方法包括使用多个特征检测器检测输入图案的特征，基于其输出值选择特征检测器中的至少一个，并且基于来自至少一个所选择的输入值的输出值来计算输入图案的特征量 特征检测器

公开(公告)号：US20080181508A1

公开(公告)日：2008-07-31

申请号：US12021659

申请日：2008-01-29

申请人： Yuji Kaneda ,  Masakazu Matsugu ,  Katsuhiko Mori

发明人： Yuji Kaneda ,  Masakazu Matsugu ,  Katsuhiko Mori

IPC分类号： G06K9/46

CPC分类号： G06K9/00281 ,  G06K9/00248 ,  G06K9/00308

摘要： A feature point detection unit (153) and feature amount extraction unit (154) extract a plurality of features of an object from input image data. When there are unextracted features of the plurality of features, a weight setting unit (155) sets weights for the extracted features. A facial expression determination unit (156) executes recognition processing of the object based on the features weighted by the weight setting unit (155).

摘要翻译： 特征点检测单元（153）和特征量提取单元（154）从输入图像数据中提取对象的多个特征。 当存在多个特征的未提取的特征时，权重设置单元（155）设置所提取的特征的权重。 面部表情确定单元（156）基于由权重设定单元（155）加权的特征来执行对象的识别处理。

公开(公告)号：US20070270664A1

公开(公告)日：2007-11-22

申请号：US11665862

申请日：2005-10-25

申请人： Mie Ishii ,  Masakazu Matsugu ,  Katsuhiko Mori ,  Yusuke Mitarai

发明人： Mie Ishii ,  Masakazu Matsugu ,  Katsuhiko Mori ,  Yusuke Mitarai

IPC分类号： A61B5/16

CPC分类号： G06F19/3418 ,  A61B5/167 ,  G06F19/00 ,  G16H50/20 ,  Y10S128/92

摘要： A condition analysis unit (1302) estimates a surrounding environment and psychological state of the user on the basis of image data, voice data, and living body information. When the estimated psychological state is a predetermined state, a cause estimation unit (1303) estimates based on the living body information whether or not the physical condition of the user is bad. When the estimated psychological state is the predetermined state, the cause estimation unit estimates a cause of the psychological state on the basis of the surrounding environment.

摘要翻译： 条件分析单元（1302）基于图像数据，语音数据和生物体信息估计用户的周围环境和心理状态。 当估计心理状态是预定状态时，原因估计单元（1303）基于生物体信息估计用户的身体状况是否不良。 当估计心理状态是预定状态时，原因估计单元基于周围环境来估计心理状态的原因。

公开(公告)号：US20070244842A1

公开(公告)日：2007-10-18

申请号：US11579981

申请日：2005-06-02

申请人： Mie Ishii ,  Masakazu Matsugu ,  Katsuhiko Mori ,  Yusuke Mitarai

发明人： Mie Ishii ,  Masakazu Matsugu ,  Katsuhiko Mori ,  Yusuke Mitarai

IPC分类号： G06F15/18

CPC分类号： G06K9/4628 ,  G06N3/04 ,  G06N3/082 ,  G06T7/12 ,  G06T2207/20084 ,  G06T2207/30201

摘要： An output value of neuron within an objective layer of a hierarchical neural network is computed. The data of the output value of neuron is stored in a memory only if the output value of neuron is greater than or equal to a predetermined value by referring to the computed output value of neuron within the objective layer. When the data of the output value of neuron on a former layer of objective layer is read from the memory, the data having a predetermined value is read, instead of the data of the output value of neuron not stored in the memory.

摘要翻译： 计算层次神经网络目标层内神经元的输出值。 如果神经元的输出值大于或等于预定值，则通过参照计算出的目标层内神经元的输出值，将神经元的输出值的数据存储在存储器中。 当从存储器中读取目标层之前的神经元的输出值的数据时，读取具有预定值的数据，而不是未存储在存储器中的神经元的输出值的数据。

公开(公告)号：US20070162828A1

公开(公告)日：2007-07-12

申请号：US10580327

申请日：2005-01-24

申请人： Yusuke Mitarai ,  Masakazu Matsugu ,  Katsuhiko Mori ,  Takashi Morie

发明人： Yusuke Mitarai ,  Masakazu Matsugu ,  Katsuhiko Mori ,  Takashi Morie

IPC分类号： H03M13/00

CPC分类号： H04N19/192 ,  H04N19/119 ,  H04N19/136 ,  H04N19/172 ,  H04N19/50 ,  H04N19/60

摘要： Data in multidimensional space such as a two-dimensional image is encoded with high efficiency. Further, as two-dimensional data can be decomposed to one-dimensional bases, the problem of wiring for two-dimensional parallelizing in a convolution arithmetic unit can be solved. For this purpose, two-dimensional image data f(x,y) to be encoded is inputted, and one-dimensional adaptive bases X(x),Y(y) representing the two-dimensional image are obtained. Next, a reconstructed image is generated based on the one-dimensional adaptive bases, and the one-dimensional adaptive bases are corrected based on an error E between the reconstructed image and the input image. The correction is repeated until the error E is reduced.

摘要翻译： 多维空间中的数据（如二维图像）以高效率进行编码。 此外，由于二维数据可以分解为一维的基础，因此可以解决卷积运算单元中二维并行化的布线问题。 为此，输入要编码的二维图像数据f（x，y），并且获得表示二维图像的一维自适应基准X（x），Y（y）。 接下来，基于一维自适应基础生成重建图像，并且基于重构图像和输入图像之间的误差E来校正一维自适应基准。 重复校正直到误差E减小。