公开(公告)号：WO2015158887A3

公开(公告)日：2016-01-21

申请号：PCT/EP2015058373

申请日：2015-04-17

Applicant: ALDEBARAN ROBOTICS

Inventor： MONCEAUX JÉRÔME ,  GATE GWENNAEL ,  HOUSSIN DAVID ,  BARBIERI GABRIELE ,  MARTIN JOCELYN ,  TESTARD JEAN ,  GOURDIN ILMO

IPC: B25J9/16 ,  B25J11/00 ,  G06N3/00 ,  G10L15/32

CPC classification number: G10L15/1815 ,  B25J9/1694 ,  B25J11/0015 ,  B25J11/005 ,  B25J13/003 ,  G06F17/271 ,  G06N3/008 ,  G10L15/18 ,  G10L15/22 ,  G10L15/32 ,  G10L2015/025

Abstract: A method of performing a dialogue between a humanoid robot (R) and at least one user (U) comprising the following steps, carried out iteratively by said humanoid robot: i) acquiring a plurality of input signals (s1, s2) from respective sensors (c1, c2), at least one said sensor being a sound sensor and at least one other sensor being a motion or image sensor; ii) interpreting the acquired signals to recognize a plurality of events (EVI) generated by said user, selected from a group comprising: the utterance of at least a word or sentence, an intonation of voice, a gesture, a body posture, a facial expression; iii) determining a response of said humanoid robot, comprising at least one event (EVO) selected from a group comprising: the utterance of at least a word or sentence, an intonation of voice, a gesture, a body posture, a facial expression; iv ) generating, by said humanoid robot, said or each said event; characterized in that said step iii) comprises determining said response as a function of at least two events jointly generated by said user and recognized at said step ii), of which at least one is not a word or sentence uttered by said user. A computer program product and a humanoid robot for carrying out such a method.

Abstract translation: 一种执行人形机器人（R）与包括以下步骤的至少一个用户（U）之间的对话的方法，由所述人形机器人迭代地进行：i）从各个传感器获取多个输入信号（s1，s2） （c1，c2），至少一个所述传感器是声音传感器，并且至少一个其它传感器是运动或图像传感器; ii）解释获取的信号以识别由所述用户生成的多个事件（EVI），其从包括以下的组中选择：至少一个单词或句子的发音，语音的语调，手势，身体姿势，面部 表达; iii）确定所述人形机器人的响应，包括从包括以下组的至少一个事件（EVO）中选择：至少一个单词或句子的发音，语音的语调，手势，身体姿势，面部表情; iv）由所述类人机器人产生所述或每个所述事件; 其特征在于，所述步骤iii）包括将所述响应确定为由所述用户共同生成并在所述步骤ii）识别的至少两个事件的函数，其中至少一个不是所述用户发出的单词或句子。 一种用于执行这种方法的计算机程序产品和人形机器人。