公开(公告)号：US11544563B2

公开(公告)日：2023-01-03

申请号：US16905539

申请日：2020-06-18

Applicant: OLYMPUS CORPORATION

Inventor： Jun Ando

IPC: G06N3/04 ,  G06N3/08

Abstract: A data processing device applies a first convolutional neural network layer to pieces of data included in a mini-batch to obtain a first feature map of each of the pieces of data, independently calculates a first statistic for each of the pieces of data based on the first feature maps, calculates a normalization parameter for each of the pieces of data based on the first statistic of each of the pieces of data and a cumulative statistic, normalizes the first feature map of each of the pieces of data by using a normalization parameter of each of the pieces of data to obtain a normalized feature map of each of the pieces of data, and applies a second convolutional neural network layer to the normalized feature map of each of the pieces of data to obtain a second feature map of each of the pieces of data.

公开(公告)号：US10089548B2

公开(公告)日：2018-10-02

申请号：US15258343

申请日：2016-09-07

Applicant: OLYMPUS CORPORATION

Inventor： Jun Ando

IPC: G06K9/32 ,  G06K9/62 ,  H04N5/345 ,  H04N5/343 ,  H04N5/347 ,  G06K9/46

Abstract: An image recognition device includes: a plurality of first charge storage circuits that store signal charges generated by photoelectric conversion sections; a plurality of second charge storage circuits that store signal charges generated by the photoelectric conversion sections; a first charge read circuit section that reads a pixel signal and outputs an image as a first image; a second charge read circuit section that reads a pixel signal and outputs an image as a second image; a read circuit selection section that selects one of the first charge read circuit section and the second charge read circuit section; and a feature amount determination section, wherein the feature amount determination section determines a detection target subject according to a feature amount of a subject in the second image, and whether to perform the determination for a subject in the first image is determined based on the determination result.

公开(公告)号：US12026935B2

公开(公告)日：2024-07-02

申请号：US17569205

申请日：2022-01-05

Applicant: OLYMPUS CORPORATION

Inventor： Jun Ando

IPC: G06V10/774 ,  G06N3/045 ,  G06V10/764 ,  G06V10/776 ,  G06V10/82 ,  G06V10/98 ,  G16H30/40

CPC classification number: G06V10/774 ,  G06N3/045 ,  G06V10/764 ,  G06V10/776 ,  G06V10/82 ,  G06V10/98 ,  G16H30/40 ,  G06V2201/032

Abstract: An image processing method includes generating a first augmented image by applying first data augmentation on an input image, generating a second augmented image by applying second data augmentation on the input image, generating a first output by inputting the first augmented image to a neural network, generating a second output by inputting the second augmented image to the neural network, calculating an output difference indicating a degree of difference between the first output and the second output, and updating a weight coefficient of each layer of the neural network based on the output difference.

公开(公告)号：US09552536B2

公开(公告)日：2017-01-24

申请号：US14878210

申请日：2015-10-08

Applicant: OLYMPUS CORPORATION

Inventor： Jun Ando

IPC: G06K9/62 ,  G06K9/46 ,  G06K9/72

CPC classification number: G06K9/6262 ,  G06K9/4642 ,  G06K9/4671 ,  G06K9/4676 ,  G06K9/6256 ,  G06K9/72

Abstract: An image processing device includes an input reception section that receives a learning image and a correct answer label, a processing section that performs a process that generates classifier data and a processing target image, and a storage section. The processing section generates the processing target image that is the entirety or part of the learning image, calculates a feature quantity of the processing target image, generates the classifier data based on training data that is a set of the feature quantity and the correct answer label assigned to the learning image that corresponds to the feature quantity, generates an image group based on the learning image or the processing target image, classifies each image of the image group using the classifier data to calculate a classification score of each image, and regenerates the processing target image based on the classification score and the image group.

Abstract translation: 图像处理装置包括接收学习图像和正确答案标签的输入接收部分，执行生成分类数据和处理目标图像的处理的处理部分以及存储部分。 处理部生成作为学习图像的全部或一部分的处理对象图像，计算处理对象图像的特征量，基于作为特征量的集合和正确答案标签的训练数据生成分类数据 分配给对应于特征量的学习图像，基于学习图像或处理对象图像生成图像组，使用分类器数据对图像组的每个图像进行分类，以计算每个图像的分类分数，并且再生 基于分类分数和图像组处理目标图像。

公开(公告)号：US20160379075A1

公开(公告)日：2016-12-29

申请号：US15258343

申请日：2016-09-07

Applicant: OLYMPUS CORPORATION

Inventor： Jun Ando

IPC: G06K9/32 ,  G06K9/46 ,  G06K9/62 ,  H04N5/347

CPC classification number: G06K9/3241 ,  G06K9/4642 ,  G06K9/6267 ,  H04N5/343 ,  H04N5/345 ,  H04N5/347

Abstract: An image recognition device includes: a plurality of first charge storage circuits that store signal charges generated by photoelectric conversion sections; a plurality of second charge storage circuits that store signal charges generated by the photoelectric conversion sections; a first charge read circuit section that reads a pixel signal and outputs an image as a first image; a second charge read circuit section that reads a pixel signal and outputs an image as a second image; a read circuit selection section that selects one of the first charge read circuit section and the second charge read circuit section; and a feature amount determination section, wherein the feature amount determination section determines a detection target subject according to a feature amount of a subject in the second image, and whether to perform the determination for a subject in the first image is determined based on the determination result.

Abstract translation: 图像识别装置包括：多个第一电荷存储电路，存储由光电转换部产生的信号电荷; 多个第二电荷存储电路，存储由所述光电转换部生成的信号电荷; 第一电荷读取电路部分，其读取像素信号并输出​​图像作为第一图像; 第二电荷读取电路部分，其读取像素信号并输出​​图像作为第二图像; 读取电路选择部分，其选择第一充电读取电路部分和第二充电读取电路部分中的一个; 特征量确定部，其特征量确定部根据第二图像中的被摄体的特征量确定检测目标被摄体，并且基于该确定确定是否执行第一图像中的被摄体的确定 结果。

公开(公告)号：US11721086B2

公开(公告)日：2023-08-08

申请号：US17174391

申请日：2021-02-12

Applicant: OLYMPUS CORPORATION

Inventor： Jun Ando

IPC: G06V10/75 ,  G06T3/40 ,  G06T5/00 ,  G06V10/764 ,  G06V10/774 ,  G06V10/82

CPC classification number: G06V10/758 ,  G06T3/40 ,  G06T5/002 ,  G06T5/008 ,  G06V10/764 ,  G06V10/7747 ,  G06V10/82 ,  G06T2207/20081 ,  G06T2207/30004

Abstract: An image processing system includes a memory storing a training image set and a reference image set, and a processor including hardware. The processor is configured to: generate an augmented image set by applying data augmentation to images included in the training image set; and determine an augmentation parameter based on a similarity between an augmentation feature statistic and a reference feature statistic, the augmentation feature statistic being a statistic of a feature of a recognition target calculated based on the augmented image set, the reference feature statistic being a statistic of a feature of the recognition target calculated based on the reference image set.