公开(公告)号：US20230237845A1

公开(公告)日：2023-07-27

申请号：US18119342

申请日：2023-03-09

Applicant: FUJITSU LIMITED

Inventor： JUNYA SAITO ,  Akiyoshi UCHIDA ,  Kentaro MURASE

IPC: G06V40/16 ,  G06T7/20 ,  G06V10/44 ,  G06V10/70 ,  G06V20/70

CPC classification number: G06V40/176 ,  G06T7/20 ,  G06V10/44 ,  G06V10/70 ,  G06V20/70 ,  G06T2207/20081 ,  G06T2207/30201

Abstract: A computer-readable recording medium has stored a program that causes a computer to execute a process including: generating a trained model that includes performing machine learning of a 1st_model based on a 1st_output value that is obtained when a 1st_image is input to the 1st_model in response to input of training data containing pair of the 1st_image and a 2nd_image and containing a 1st_label indicating which of the 1st and 2nd_image has captured greater movement of muscles of facial expression of a photographic subject, a 2nd_output value obtained when the 2nd_image is input to a 2nd_model that has common parameters with the 1st_model, and the 1st_label; and generating a 3rd_model that includes performing machine learning based on a 3rd_output value obtained when a 3rd_image is input to the trained model, and a 2nd_label indicating of movement of muscles of facial expression of a photographic subject captured in the 3rd_image.

公开(公告)号：US20210248357A1

公开(公告)日：2021-08-12

申请号：US17157602

申请日：2021-01-25

Applicant: FUJITSU LIMITED

Inventor： Ryosuke KAWAMURA ,  Kentaro MURASE

IPC: G06K9/00 ,  G06N20/00

Abstract: An image processing method executed by a computer, the method includes detecting a plurality of feature points of a face from an input image, referring to importance information that indicates an importance of a region within an image in a process of detecting a predetermined facial motion from the image, selecting, from the plurality of feature points detected by the detecting, one or more points that correspond to an image region including an importance indicated by the importance information equal to or smaller than a first threshold value, correcting the input image by using the one or more points selected by the selecting, to generate a corrected image; and determining whether or not the predetermined facial motion is occurring in the input image, based on an output obtained by inputting the corrected image to a recognition model.

公开(公告)号：US20220415085A1

公开(公告)日：2022-12-29

申请号：US17718344

申请日：2022-04-12

Applicant: FUJITSU LIMITED

Inventor： Ryosuke KAWAMURA ,  Kentaro MURASE

IPC: G06V40/16 ,  G06T7/20 ,  G06T7/00 ,  G06N20/20

Abstract: A non-transitory computer-readable recording medium stores a program that causes a computer to execute a process, the process includes inputting each of first images that includes a face of a subject to a first machine learning model to obtain a recognition result that includes information indicating first occurrence probability of each of facial expressions in each first image, generating training data that includes the recognition result and second images that are respectively generated based on the first images and in which at least a part of the face of the subject is concealed, and performing training of a second machine learning model, based on the training data, by using a loss function that represents an error that relates to a second occurrence probability of each facial expression in each second image and relates to magnitude relationship in the second occurrence probability among the second images.

公开(公告)号：US20240212352A1

公开(公告)日：2024-06-27

申请号：US18472672

申请日：2023-09-22

Applicant: Fujitsu Limited

Inventor： Yuka JO ,  Kentaro MURASE

IPC: G06V20/52 ,  G06T7/20 ,  G06V10/25 ,  G06V10/764 ,  G06V40/20

CPC classification number: G06V20/52 ,  G06T7/20 ,  G06V10/25 ,  G06V10/764 ,  G06V40/20 ,  G06T2207/30196 ,  G06V2201/07

Abstract: A non-transitory computer-readable recording medium stores an information output program for causing a computer to execute processing including: acquiring a video of an inside of a store; specifying, by analyzing the acquired video, a first area that includes a store clerk, a second area that includes an object customer who purchases a product, and a first relationship that identifies an interaction between the store clerk and the customer, from the video; determining that the store clerk included in the first area has performed service behavior with respect to the customer included in the second area based on the specified first relationship; and outputting a result of the determination to a display device.

公开(公告)号：US20160275050A1

公开(公告)日：2016-09-22

申请号：US15057826

申请日：2016-03-01

Applicant: FUJITSU LIMITED

Inventor： Masakiyo TANAKA ,  Jun TAKAHASHI ,  Kentaro MURASE

IPC: G06F17/21 ,  G06F17/27 ,  G10L15/22

CPC classification number: G10L15/22 ,  G06F17/2229 ,  G06F17/2235 ,  G10L15/26 ,  G10L25/48 ,  G10L2015/088 ,  H04N21/85

Abstract: A presentation supporting device extracts a first word from a character string included in each region divided from a page of a document file, and calculates a score, for each region in a page currently-displayed, based on the first word and a second word acquired as a result of a sound recognition, and calculates, when the highest score of scores is equal to or higher than a first threshold, a distance between a first region in which the highlight display is currently executed and a second region in which the highest score is equal to or higher than the first threshold, and executes a highlight display in the second region when a frequency corresponding to the distance between the first region and the second region is equal to or higher than a second threshold, and executes a highlight display in the first region, when the second threshold is not reached.

Abstract translation: 呈现支持装置从包含在从文件文件的页面分割的每个区域中的字符串中提取第一个单词，并且基于第一个单词和第二个单词计算当前显示的页面中每个区域的分数 作为声音识别的结果，并且当得分的最高得分等于或高于第一阈值时，计算当前执行高亮显示的第一区域与其中最高分数的第二区域之间的距离 等于或高于第一阈值，并且当与第一区域和第二区域之间的距离相对应的频率等于或高于第二阈值时，在第二区域中执行高光显示，并且执行高亮显示 第一个区域，当没有达到第二个阈值时。

公开(公告)号：US20250131767A1

公开(公告)日：2025-04-24

申请号：US19000740

申请日：2024-12-24

Applicant: FUJITSU LIMITED

Inventor： Junya SAITO ,  Ryosuke KAWAMURA ,  Sachihiro YOUOKU ,  Akiyoshi UCHIDA ,  Xiaoyu MI ,  Kentaro MURASE

IPC: G06V40/16

Abstract: A non-transitory computer-readable recording medium storing an information processing program for causing a computer to execute processing includes acquiring a first face image, specifying a first state of elements of an imaging condition from the first face image, generating a second state of the elements of the imaging condition changed such that the first state is improved, inputting the second state to a machine learning model for each of action units (AUs) that represent movements of facial expression muscles, with states of the elements of the imaging condition as features and errors in estimated values with respect to ground truth values of intensities of the AUs as ground truth data, to estimate prediction errors for each of the Aus, determining whether or not predetermined criteria are satisfied by all of the prediction errors, and specifying the elements of the imaging condition suitable to be improved on the first face image.

公开(公告)号：US20240037986A1

公开(公告)日：2024-02-01

申请号：US18308179

申请日：2023-04-27

Applicant: Fujitsu Limited

Inventor： Ryosuke KAWAMURA ,  Kentaro MURASE

IPC: G06V40/16 ,  G06V10/774 ,  G06V10/26

CPC classification number: G06V40/176 ,  G06V40/172 ,  G06V40/168 ,  G06V10/774 ,  G06V10/26

Abstract: A recording medium stores a program for causing a computer to execute processing including: acquiring images; classifying the images, based on a combination of whether an action unit related to a motion of a portion occurs and whether occlusion is included in an image in which the action unit occurs; calculating a feature amount of the image by inputting each classified image into a model; and training the model so as to decrease a first distance between feature amounts of an image in which the action unit occurs and an image with an occlusion with respect to the image in which the action unit occurs and to increase a second distance between feature amounts of the image with the occlusion with respect to the image in which the action unit occurs and an image with an occlusion with respect to an image in which the action unit does not occur.