公开(公告)号：US20160371254A1

公开(公告)日：2016-12-22

申请号：US15176114

申请日：2016-06-07

Applicant: Panasonic Intellectual Property Management Co., Ltd.

Inventor： KATSUYOSHI YAMAGAMI ,  TAKASHI USHIO ,  YASUNORI ISHII

IPC: G06F17/27 ,  G06N3/08 ,  G06N3/04 ,  G06F19/00

CPC classification number: G06F17/2785 ,  G06F17/2795 ,  G06F19/00 ,  G06F19/36 ,  G06N3/0472 ,  G06N3/08 ,  G16H10/20

Abstract: A method includes acquiring a first corpus, including first text of a first sentence including a first word and described in a natural language, and second text of a second sentence including a second word different in meaning from the first word, a second word distribution of the second word being similar to a first word distribution of the first word, acquiring a second corpus including third text of a third sentence, including a third word identical to the first word and/or the second word, a third word distribution of the third word being not similar to the first word distribution, based on an arrangement of a word string in the first corpus and the second corpus, assigning to the first word a first vector representing a meaning of the first word and assigning to the second word a second vector representing a meaning of the second word.

Abstract translation: 一种方法，包括：获取第一语料库，包括包括第一单词并以自然语言描述的第一句子的第一文本，以及第二句子的第二文本，其包括与第一单词不同意义的第二单词，第二单词分布 所述第二词类似于所述第一单词的第一单词分布，获取包括第三句子的第三文本的第二语料库，包括与所述第一单词和/或所述第二单词相同的第三单词，所述第三单词分布 基于第一语料库和第二语料库中的单词串的排列，将第一单词与第一单词分布不相似，向第一单词分配表示第一单词的含义的第一向量，并将第二单词分配给第二单词 代表第二个词的含义的向量。

公开(公告)号：US20160260014A1

公开(公告)日：2016-09-08

申请号：US15053642

申请日：2016-02-25

Applicant: Panasonic Intellectual Property Management Co., Ltd.

Inventor： REIKO HAGAWA ,  SOTARO TSUKIZAWA ,  YASUNORI ISHII

IPC: G06N3/08 ,  G06K9/62

CPC classification number: G06N3/08 ,  G06K7/1482 ,  G06K9/4628 ,  G06K9/6267 ,  G06K9/6281 ,  G06K9/6282 ,  G06N3/02 ,  G06N3/0454 ,  G06T2207/20081

Abstract: Learning method includes performing a first process in which a coarse class classifier configured with a first neural network is made to classify a plurality of images given as a set of images each attached with a label indicating a detailed class into a plurality of coarse classes including a plurality of detailed classes and is then made to learn a first feature that is a feature common in each of the coarse classes, and performing a second process in which a detailed class classifier, configured with a second neural network that is the same in terms of layers other than the final layer as but different in terms of the final layer from the first neural network made to perform the learning in the first process, is made to classify the set of images into detailed classes and learn a second feature of each detailed class.

Abstract translation: 学习方法包括执行第一处理，其中配置有第一神经网络的粗类分类器被分类为以图像集合给出的多个图像，每个图像附加有指示详细类别的标签为多个粗类，包括 多个详细类，然后被做成学习作为每个粗类中共同的特征的第一特征，并且执行第二处理，其中配置有第二神经网络的详细类分类器在第 将最终层以外的层作为第一神经网络的最终层而不是第一神经网络，以在第一过程中执行学习进行分类，以将该组图像分类为详细类并学习每个详细类的第二特征 。

公开(公告)号：US20160259995A1

公开(公告)日：2016-09-08

申请号：US15049149

申请日：2016-02-22

Applicant: Panasonic Intellectual Property Management Co., Ltd.

Inventor： YASUNORI ISHII ,  SOTARO TSUKIZAWA ,  REIKO HAGAWA

IPC: G06K9/62 ,  G06N3/08 ,  G06K9/46

CPC classification number: G06K9/6267 ,  G06K9/46 ,  G06K9/4628 ,  G06K9/6256 ,  G06K9/6274 ,  G06N3/0454 ,  G06N3/08

Abstract: An image recognition method includes: receiving an image; acquiring processing result information including values of processing results of convolution processing at positions of a plurality of pixels that constitute the image by performing the convolution processing on the image by using different convolution filters; determining 1 feature for each of the positions of the plurality of pixels on the basis of the values of the processing results of the convolution processing at the positions of the plurality of pixels included in the processing result information and outputting the determined feature for each of the positions of the plurality of pixels; performing recognition processing on the basis of the determined feature for each of the positions of the plurality of pixels; and outputting recognition processing result information obtained by performing the recognition processing.

Abstract translation: 一种图像识别方法，包括：接收图像; 通过使用不同的卷积滤波器对所述图像执行卷积处理来获取包括在构成图像的多个像素的位置处的卷积处理的处理结果的值的处理结果信息; 基于处理结果信息中包括的多个像素的位置处的卷积处理的处理结果的值，确定多个像素中的每个位置的1个特征，并输出所确定的特征 所述多个像素的位置; 基于针对所述多个像素的每个位置的所确定的特征执行识别处理; 并输出通过执行识别处理获得的识别处理结果信息。