公开(公告)号：US20110010176A1

公开(公告)日：2011-01-13

申请号：US12829984

申请日：2010-07-02

申请人： Yukiko YOSHIIKE ,  Kenta Kawamoto ,  Kuniaki Noda ,  Kohtaro Sabe

发明人： Yukiko YOSHIIKE ,  Kenta Kawamoto ,  Kuniaki Noda ,  Kohtaro Sabe

IPC分类号： G10L15/14

CPC分类号： G06N99/005

摘要： An HMM (Hidden Markov Model) learning device includes: a learning unit for learning a state transition probability as the function of actions that an agent can execute, with learning with HMM performed based on actions that the agent has executed, and time series information made up of an observation signal; and a storage unit for storing learning results by the learning unit as internal model data including a state-transition probability table and an observation probability table; with the learning unit calculating frequency variables used for estimation calculation of HMM state-transition and HMM observation probabilities; with the storage unit holding the frequency variables corresponding to each of state-transition probabilities and each of observation probabilities respectively, of the state-transition probability table; and with the learning unit using the frequency variables held by the storage unit to perform learning, and estimating the state-transition probability and the observation probability based on the frequency variables.

摘要翻译： HMM（隐马尔可夫模型）学习装置包括：用于学习作为代理可以执行的动作的函数的状态转移概率的学习单元，基于代理已经执行的动作执行的HMM的学习以及作出的时间序列信息 观察信号; 以及存储单元，用于将所述学习单元的学习结果存储为包括状态转换概率表和观察概率表的内部模型数据; 学习单元计算用于HMM状态转换和HMM观察概率的估计计算的频率变量; 存储单元保持状态转移概率表中分别对应于状态转换概率和每个观察概率的频率变量; 并且所述学习单元使用由所述存储单元保持的频率变量来执行学习，并且基于所述频率变量来估计所述状态转换概率和观察概率。

公开(公告)号：US20100318478A1

公开(公告)日：2010-12-16

申请号：US12791240

申请日：2010-06-01

申请人： Yukiko Yoshiike ,  Kenta Kawamoto ,  Kuniaki Noda ,  Kohtaro Sabe

发明人： Yukiko Yoshiike ,  Kenta Kawamoto ,  Kuniaki Noda ,  Kohtaro Sabe

IPC分类号： G06N5/02 ,  G06F15/18

CPC分类号： G06N3/006 ,  G06K9/00335 ,  G06K9/00664 ,  G06K9/6297

摘要： An information processing device includes: a calculating unit configured to calculate a current-state series candidate that is a state series for an agent capable of actions reaching the current state, based on a state transition probability model obtained by performing learning of the state transition probability model stipulated by a state transition probability that a state will be transitioned according to each of actions performed by an agent capable of actions, and an observation probability that a predetermined observation value will be observed from the state, using an action performed by the agent, and an observation value observed at the agent when the agent performs an action; and a determining unit configured to determine an action to be performed next by the agent using the current-state series candidate in accordance with a predetermined strategy.

摘要翻译： 一种信息处理设备，包括：计算单元，被配置为基于通过执行状态转移概率的学习获得的状态转移概率模型来计算作为达到当前状态的动作的代理的状态序列的当前状态序列候选 由状态转移概率规定的模式，其状态将根据由能够执行动作的代理执行的动作而转变，并且使用由代理执行的动作从该状态观察到预定观察值的观察概率， 以及当代理人执行动作时在代理处观察到的观察值; 以及确定单元，被配置为根据预定策略来确定所述代理使用所述当前状态序列候选来执行的动作。

公开(公告)号：US07751937B2

公开(公告)日：2010-07-06

申请号：US11350825

申请日：2006-02-10

申请人： Kohtaro Sabe ,  Kenichi Hidai

发明人： Kohtaro Sabe ,  Kenichi Hidai

IPC分类号： G06F19/00

CPC分类号： G06N99/005 ,  G05B13/021 ,  G05B13/026 ,  G05B13/0265 ,  G05B13/027

摘要： A behavior control apparatus to control behavior of a device capable of sensing a state of an environment and selecting an action on the basis of a sensing result is provided. The behavior control apparatus includes a predicting unit configured to learn the action and change in the state of the environment and predict change in the state of the environment caused by a predetermined action on the basis of the learning; a planning unit configured to plan a behavior sequence to achieve a goal state from a present state on the basis of the prediction made by the predicting unit; and a control unit configured to control each action of the behavior sequence planned by the planning unit and learn an input/output relationship if the goal state is achieved through the action.

摘要翻译： 提供了一种用于控制能够感测环境状态并且基于感测结果选择动作的装置的行为的行为控制装置。 行为控制装置包括：预测单元，被配置为基于学习来学习环境状态的动作和变化，并且预测由预定动作引起的环境状态的变化; 计划单元，被配置为基于由所述预测单元进行的预测来规划行为序列以从当前状态实现目标状态; 以及控制单元，被配置为控制由规划单元计划的行为序列的每个动作，并且如果通过动作实现目标状态，则学习输入/输出关系。

公开(公告)号：US07725412B2

公开(公告)日：2010-05-25

申请号：US11732644

申请日：2007-04-04

申请人： Kazumi Aoyama ,  Kohtaro Sabe ,  Hideki Shimomura

发明人： Kazumi Aoyama ,  Kohtaro Sabe ,  Hideki Shimomura

IPC分类号： G06F15/00

CPC分类号： G06F17/30551 ,  G06K9/0057 ,  G06K9/6251

摘要： A data processing device for processing time-sequence data includes a data extracting unit extracting time-sequence data for a predetermined time unit from time-sequence data; and a processing unit obtaining scores for nodes of an SOM configured from multiple nodes provided with a spatial array configuration, the scores showing applicability to time-sequence data for a predetermined time unit thereof. The node with the best score is determined to be the winning node which is the node most applicable. The processing unit obtains scores as to the time-sequence data for one predetermined time unit, regarding a distance-restricted node wherein distance from the winning node as to the time-sequence for a predetermined time unit immediately preceding the time-sequence data of one predetermined time unit is within a predetermined distance. The distance-restricted node with the best the score is determined to be the winning node.

摘要翻译： 一种用于处理时间序列数据的数据处理装置，包括数据提取单元，从时间序列数据提取预定时间单位的时间序列数据; 以及处理单元，其获得由具有空间阵列配置的多个节点配置的SOM的节点，所述分数显示对于其预定时间单位的时间序列数据的适用性。 具有最佳分数的节点被确定为作为最适用节点的获胜节点。 处理单元获得关于一个预定时间单位的时间序列数据的分数，关于距离限制节点的距离限制节点，其中距离胜利节点的距离与紧接在一个的时间序列数据之前的预定时间单位的时间序列 预定时间单位在预定距离内。 得分最高的距离限制节点被确定为获胜节点。

公开(公告)号：US07587069B2

公开(公告)日：2009-09-08

申请号：US12075080

申请日：2008-03-07

申请人： Javier R. Movellan ,  Marian S. Bartlett ,  Gwendolen C. Littlewort ,  John Hershey ,  Ian R. Fasel ,  Eric C. Carlson ,  Josh Susskind ,  Kohtaro Sabe ,  Kenta Kawamoto ,  Kenichi Hidai

发明人： Javier R. Movellan ,  Marian S. Bartlett ,  Gwendolen C. Littlewort ,  John Hershey ,  Ian R. Fasel ,  Eric C. Carlson ,  Josh Susskind ,  Kohtaro Sabe ,  Kenta Kawamoto ,  Kenichi Hidai

IPC分类号： G06K9/00

CPC分类号： G06K9/00248 ,  G06K9/00308 ,  G06K9/6256 ,  G06N3/004

摘要： A facial expression recognition system that uses a face detection apparatus realizing efficient learning and high-speed detection processing based on ensemble learning when detecting an area representing a detection target and that is robust against shifts of face position included in images and capable of highly accurate expression recognition, and a learning method for the system, are provided. When learning data to be used by the face detection apparatus by Adaboost, processing to select high-performance weak hypotheses from all weak hypotheses, then generate new weak hypotheses from these high-performance weak hypotheses on the basis of statistical characteristics, and select one weak hypothesis having the highest discrimination performance from these weak hypotheses, is repeated to sequentially generate a weak hypothesis, and a final hypothesis is thus acquired. In detection, using an abort threshold value that has been learned in advance, whether provided data can be obviously judged as a non-face is determined every time one weak hypothesis outputs the result of discrimination. If it can be judged so, processing is aborted. A predetermined Gabor filter is selected from the detected face image by an Adaboost technique, and a support vector for only a feature quantity extracted by the selected filter is learned, thus performing expression recognition.

公开(公告)号：US20090175533A1

公开(公告)日：2009-07-09

申请号：US12366792

申请日：2009-02-06

申请人： Kohtaro Sabe ,  Kenichi Hidai

发明人： Kohtaro Sabe ,  Kenichi Hidai

IPC分类号： G06K9/62

CPC分类号： G06K9/6256 ,  G06K9/00248 ,  G06K9/6282

摘要： In an information processing apparatus, such as a robot that discriminates human faces, nodes are hierarchically arranged in a tree structure. Each of the nodes has a number of weak classifiers. Each terminal node learns face images associated with one label. An upper node learns learning samples of all labels learned by lower nodes. When a window image to be classified is input, discrimination is performed sequentially from upper nodes to lower nodes. When it is determined that the window image does not correspond to a human face, discrimination by lower nodes is not performed, and discrimination proceeds to sibling nodes.

摘要翻译： 在诸如识别人脸的机器人等信息处理装置中，以树结构分层地布置节点。 每个节点都有一些弱分类器。 每个终端节点学习与一个标签相关联的脸部图像。 上级节点学习下层节点学习的所有标签的学习样本。 当输入要分类的窗口图像时，从上层节点到下层节点依次进行鉴别。 当确定窗口图像不对应于人脸时，不执行下位节点的鉴别，并且歧视进行到兄弟节点。

公开(公告)号：US07369686B2

公开(公告)日：2008-05-06

申请号：US10399740

申请日：2002-08-21

申请人： Jun Yokono ,  Kohtaro Sabe ,  Kenta Kawamoto

发明人： Jun Yokono ,  Kohtaro Sabe ,  Kenta Kawamoto

IPC分类号： G06K9/00

CPC分类号： G06K9/00288 ,  G06K9/00221 ,  G06K9/00228

摘要： A robot includes a face extracting section for extracting features of a face included in an image captured by a CCD camera, and a face recognition section for recognizing the face based on a result of face extraction by the face extracting section. The face extracting section is implemented by Gabor filters that filter images using a plurality of filters that have orientation selectivity and that are associated with different frequency components. The face recognition section is implemented by a support vector machine that maps the result of face recognition to a non-linear space and that obtains a hyperplane that separates in that space to discriminate a face from a non-face. The robot is allowed to recognize a face of a user within a predetermined time under a dynamically changing environment.

摘要翻译： 机器人包括：面部提取部，用于提取由CCD照相机拍摄的图像中包含的面部的特征;以及面部识别部，其基于面部提取部的面部提取的结果识别脸部。 人脸提取部分由Gabor滤波器实现，该滤波器使用具有取向选择性并且与不同频率分量相关联的多个滤波器对图像进行滤波。 脸部识别部分由支持向量机实现，该支持向量机将人脸识别的结果映射到非线性空间，并获得在该空间中分离的超平面，以将脸部与非脸部区分开。 允许机器人在动态变化的环境下在预定时间内识别用户的脸部。

公开(公告)号：US07228201B2

公开(公告)日：2007-06-05

申请号：US10110497

申请日：2001-08-16

申请人： Kohtaro Sabe ,  Kenta Kawamoto

发明人： Kohtaro Sabe ,  Kenta Kawamoto

IPC分类号： G06F19/00

CPC分类号： G06F9/544 ,  G05B19/4187 ,  G06F9/546 ,  G06N3/008 ,  Y02P90/205

摘要： A robot device (1) has a central processing process (CPU) having a plurality of objects and adapted for carrying out control processing on the basis inter-object communication carried out between the objects, the central processing process controlling accesses by the plurality of objects to a shared memory shared by the plurality of objects and thus carrying out inter-object communication. Specifically, the central processing process generates pointers P11, P12, P13, P21, P22 in accordance with accesses by the objects to predetermined areas M1, M2 on a shared memory M, then measures the pointers by the corresponding number-of-reference measuring objects RO1, RO2, and controls the accesses in accordance with the number of pointers measured, thereby carrying out inter-object communication. This enables easy realization of smooth inter-process communication.

摘要翻译： 机器人装置（1）具有具有多个对象的中央处理处理（CPU），并且适于在对象之间进行的基于对象间通信的基础上进行控制处理，中央处理处理控制多个对象的访问 到由多个对象共享的共享存储器，从而进行对象间通信。 具体地，中央处理处理根据对象对共享存储器M上的预定区域M 1，M 2的访问来生成指针P 11，P 12，P 13，P 21，P 22，然后通过相应的 基准测量对象RO 1，RO 2，并且根据所测量的指针的数量控制访问，从而进行对象间通信。 这使得能够容易地实现平滑的进程间通信。

公开(公告)号：US07088853B2

公开(公告)日：2006-08-08

申请号：US10336201

申请日：2002-12-31

申请人： Atsuo Hiroe ,  Katsuki Minamino ,  Kenta Kawamoto ,  Kohtaro Sabe ,  Takeshi Ohashi

发明人： Atsuo Hiroe ,  Katsuki Minamino ,  Kenta Kawamoto ,  Kohtaro Sabe ,  Takeshi Ohashi

IPC分类号： G06K9/00 ,  G10L15/00 ,  G10L13/00

CPC分类号： G10L13/047 ,  G10L13/00 ,  G10L15/06 ,  G10L15/24 ,  G10L2015/0631

摘要： A plural number of letters or characters, inferred from the results of letter/character recognition of an image photographed by a CCD camera (20), a plural number of kana readings inferred from the letters or characters and the way of pronunciation corresponding to the kana readings are generated in an pronunciation information generating unit (150) and the plural readings obtained are matched to the pronunciation from the user acquired by a microphone (23) to specify one kana reading and the way of pronunciation (reading) from among the plural generated candidates.

摘要翻译： 从由CCD照相机（20）拍摄的图像的字母/字符识别的结果推断的多个字母或字符，从字母或字符推断的多个假名读数和对应于假名的发音方式 在发音信息生成单元（150）中产生读数，并且所获得的多个读数与由麦克风（23）获取的用户的发音相匹配，以从多个生成的多个生成单元中指定一个假名读数和发音（读取）的方式 候选人。

公开(公告)号：US07076334B2

公开(公告)日：2006-07-11

申请号：US10989713

申请日：2004-11-16

申请人： Kohtaro Sabe ,  Kazuo Ishii

发明人： Kohtaro Sabe ,  Kazuo Ishii

IPC分类号： G06F19/00

CPC分类号： G06N3/008 ,  A63H3/28 ,  A63H2200/00 ,  Y10S707/99953 ,  Y10S707/99955

摘要： A main robot apparatus generates a sound scale command at a command generating state (ST2) to enter into a state of waiting for a reaction of a slave robot apparatus (ST3). When the slave robot apparatus outputs a emotion expressing sound responsive to a sound scale command issued by the main robot apparatus, the main robot apparatus recognizes this emotion expressing sound to output the same emotion expressing sound. In a state of the reaction action (ST4), the main robot apparatus selects an action (NumResponse), depending on the value of the variable NumResponse which has counted the number of times of the reactions to output the action.

摘要翻译： 主机器人装置在命令生成状态（ST2）下生成声音刻度指令，进入等待从动机器人装置的反应的状态（ST3）。 当从机器人装置响应于由主机器人装置发出的声音音阶命令输出表达声音的声音时，主机器人装置识别该情绪表达声音以输出相同的表情声。 在反应动作的状态（ST4）中，主机器人装置根据已经计算了输出动作的反应次数的变量NumResponse的值来选择动作（NumResponse）。