公开(公告)号：US20230154012A1

公开(公告)日：2023-05-18

申请号：US17919584

申请日：2020-04-24

Applicant: NEC Corporation

Inventor： Shoji YACHIDA ,  Keiko INOUE ,  Azusa SAWADA ,  Toshinori HOSOI

IPC: G06T7/20 ,  G06T7/00 ,  G01N21/90

CPC classification number: G06T7/20 ,  G06T7/0002 ,  G01N21/90 ,  G06T2207/20021 ,  G06T2207/10152

Abstract: A determination apparatus includes: a dividing unit configured to divide chronological image data acquired by imaging a liquid filled in a container while switching between a plurality of illumination conditions, into chronological image data corresponding to the illumination conditions; and a determining unit configured to determine foreign matter contained in the container based on each of the chronological image data obtained by division by the dividing unit.

公开(公告)号：US20220335291A1

公开(公告)日：2022-10-20

申请号：US17640926

申请日：2019-09-20

Applicant: NEC Corporation

Inventor： Azusa SAWADA ,  Soma SHIRAISHI ,  Takashi SHIBATA

IPC: G06N3/08 ,  G06V10/771 ,  G06V10/77 ,  G06V10/82

Abstract: A learning apparatus includes a metric space learning unit and a case example storage unit. The metric space learning unit learns a metric space including feature vectors extracted from sets of attribute-attached image data for each combination of different attributes by using the sets of attribute-attached image data to which pieces of attribute information are added. The case example storage unit calculates feature vectors from sets of case example image data, and stores the feature vectors as case examples associated with the metric space.

公开(公告)号：US20220215579A1

公开(公告)日：2022-07-07

申请号：US17604494

申请日：2019-04-22

Applicant: NEC Corporation

Inventor： Azusa SAWADA ,  Takashi SHIBATA

IPC: G06T7/73

Abstract: Please delete the Abstract of the Disclosure, and replace it with the following: An input image acquisition unit acquires a plurality of input images in which a specific detection target is captured by a plurality of different modalities. A perturbed image acquisition unit acquires a plurality of perturbed images in which at least one of the plurality of input images is perturbed. A detection processing unit detects a detection target included in the input images using each of the plurality of perturbed images and one of the plurality of input images that has not been perturbed, and acquires, for each of the plurality of perturbed images, a detection position of the detection target and a detection confidence level as detection results. An adjustment unit calculates, based on the detection positions and the confidence levels acquired for the plurality of perturbed images, an adjusted confidence level for each of the perturbed images using integrated parameters.

公开(公告)号：US20220189144A1

公开(公告)日：2022-06-16

申请号：US17442757

申请日：2019-03-27

Applicant: NEC Corporation

Inventor： Hiroyoshi MIYANO ,  Masato TODA ,  Azusa SAWADA ,  Takashi SHBATA

IPC: G06V10/774 ,  G06V10/70 ,  G06N5/04

Abstract: An information processing apparatus (10) according to the present disclosure includes: an object recognition unit (11) that outputs, by using a first modal signal and a first modal recognition model corresponding to the first modal signal, an inference result regarding the first modal signal; a training data processing unit (12) that generates first modal training data regarding the first modal signal by using the inference result, and updates second modal training data regarding a second modal signal that is different from the first modal signal by using the first modal training data; and a recognition model update unit (13) that updates a second modal recognition model corresponding to the second modal signal by using the second modal training data.

公开(公告)号：US20240394340A1

公开(公告)日：2024-11-28

申请号：US18691081

申请日：2021-09-24

Applicant: NEC Corporation

Inventor： Azusa SAWADA

IPC: G06F18/2413

Abstract: A learning device includes a learning means for learning a discriminative model that discriminates a class to which second data belongs, the second data being data corresponding to an unknown object, by using first training data that includes a group including a plurality of pieces of first data corresponding to the same object, and a first data label with respect to the group. The learning means computes a discrimination score with respect to the first data by using the discriminative model, and learns the discriminative model by using a loss weighted by a weight that depends on a relative height of the discrimination score in the group.

公开(公告)号：US20240037889A1

公开(公告)日：2024-02-01

申请号：US18266343

申请日：2021-11-26

Applicant: NEC Corporation

Inventor： Kenta SENZAKI ,  Azusa SAWADA ,  Hironobu MORI ,  Kyoko MUROZONO ,  Katsuya ODAKA

IPC: G06V10/22 ,  G06V10/74 ,  G06V20/70 ,  G06V10/12 ,  G06V10/10

CPC classification number: G06V10/22 ,  G06V10/74 ,  G06V20/70 ,  G06V10/12 ,  G06V10/16

Abstract: This image processing device comprises an input unit, a verification area extraction unit, and an output unit. The input unit receives, as an annotation area, an input of information of an area on a first image in which an object subjected to annotation processing is present. The verification area extraction unit extracts a second image including the annotation area and captured in a manner different from that for the first image. The output unit outputs the first image and the second image in a comparable state.

公开(公告)号：US20230081660A1

公开(公告)日：2023-03-16

申请号：US17801849

申请日：2020-03-19

Applicant: NEC Corporation

Inventor： Azusa SAWADA ,  Takashi Shibata ,  Kazufo Takizawa

IPC: G06K9/62 ,  G06T3/40 ,  G01S13/90

Abstract: An image processing apparatus according to the present invention includes: an extracting unit configured to extract a candidate image, which is an image of a candidate region specified in accordance with a preset criterion, from a target image to be a target for an annotation process, and also extract a corresponding image, which is an image of a corresponding region corresponding to the candidate region, from a reference image that is an image corresponding to the target image; a displaying unit configured to display the candidate image and the corresponding image so as to be able to compare the images with each other; and an input accepting unit configured to accept input of input information for the annotation process for the candidate image.

公开(公告)号：US20220405534A1

公开(公告)日：2022-12-22

申请号：US17772793

申请日：2020-03-03

Applicant: NEC Corporation

Inventor： Eiji KANEKO ,  Azusa SAWADA ,  Kazutoshi SAGI

IPC: G06K9/62 ,  G06N20/00

Abstract: A prediction unit classifies input data into a plurality of classes using a predictive model, and outputs a predicted probability for each class as a prediction result. A grouping unit generates a grouped class formed by k classes within top k predicted probabilities, and calculates a predicted probability of the grouped class. A loss calculation unit calculates a loss based on predicted probabilities of a plurality of classes including the grouped class. A model update unit updates the predictive model based on the calculated loss.

公开(公告)号：US20220180195A1

公开(公告)日：2022-06-09

申请号：US17598422

申请日：2019-06-25

Applicant: NEC Corporation

Inventor： Azusa SAWADA ,  Takashi SHIBATA

IPC: G06N3/08 ,  G06N3/04 ,  G06K9/62

Abstract: The model generation device generates model parameters corresponding to the model to be used and mediation parameter relevance information indicating the relevance between the model parameters of a plurality of source domains and the mediation parameters by using the learning data in the plurality of source domains. The model adjustment device generates target model parameters which correspond to the target domain and include the mediation parameters, based on the learned model parameters for each of the plurality of source domains and the mediation parameter relevance information. Then, the model adjustment device uses the evaluation data of the target domain to determine the mediation parameters included in the target model parameters.

公开(公告)号：US20220128988A1

公开(公告)日：2022-04-28

申请号：US17431261

申请日：2019-02-19

Applicant: NEC Corporation

Inventor： Azusa SAWADA ,  Takashi SHIBATA ,  Katsuhiko TAKAHASHI

IPC: G05B23/02

Abstract: A data-series group includes data series which is a series of data obtained by observing the same object at discrete times. Time labels are time information added to respective data included in the data-series group. State labels are added to some of the data included in the data-series group. A loss-function control unit determines a loss function to be used for learning based on the time labels and the state labels. A threshold is used to adjust a branch condition of the loss-function control unit. A regressor is a model, and is used to detect an abnormality or predict a remaining life span. A dictionary stores parameters of the regressor. A regressor training unit trains the regressor based on the loss function determined by the loss-function control unit.