公开(公告)号：US20170300787A1

公开(公告)日：2017-10-19

申请号：US15483359

申请日：2017-04-10

Applicant: CANON KABUSHIKI KAISHA

Inventor： Tsewei Chen

IPC: G06K9/62 ,  G06K9/46

CPC classification number: G06K9/6267 ,  G06K9/00973 ,  G06K9/4604 ,  G06K9/6257 ,  G06K9/629

Abstract: An image processing apparatus includes a generation unit that generates feature data based on an image, classification units that classify a predetermined pattern by referring to the feature data, and a control unit that controls operations of the classification units. The classification units include a first classification unit and a second classification unit, processing results of which do not depend on each other, and a third classification unit. The first and the second classification units are operated in parallel. When either of the first and the second classification units determines that a classification condition is not satisfied, the control unit stops operations of all of the classification units. When both the first and the second classification units determine the classification condition is satisfied, the control unit operates the third classification unit by using classification results of the first and the second classification units.

公开(公告)号：US20160321521A1

公开(公告)日：2016-11-03

申请号：US15139603

申请日：2016-04-27

Applicant: CANON KABUSHIKI KAISHA

Inventor： Tsewei Chen

IPC: G06K9/62 ,  G06K9/00

CPC classification number: G06K9/6282 ,  G06K9/00228 ,  G06K9/6257 ,  G06K9/6286

Abstract: In an information processing apparatus that includes sequences of weak classifiers which are logically cascade-connected in each sequence and the sequences respectively correspond to categories of an object and in which the weak classifiers are grouped into at least a first group and a second group in the order of connection, classification processing by weak classifiers belonging to the first group of respective categories is performed by pipeline processing. Based on the processing results of the weak classifiers belonging to the first group of the respective categories, categories in which classification processing by weak classifiers belonging to the second group is to be performed are decided out of the categories. The classification processing by the weak classifiers respectively corresponding to the decided categories and belonging to the second group is performed by pipeline processing.

Abstract translation: 在包括在每个序列中逻辑级联连接的弱分类器的序列的信息处理设备中，并且所述序列分别对应于对象的类别，并且其中所述弱分类器被分组为至少第一组和 通过流水线处理执行属于第一组各自类别的弱分类器的连接顺序，分类处理。 根据属于第一组各类别的弱分类器的处理结果，根据属于第二组的弱分类器进行分类处理的类别进行判定。 分类对应于属于第二组的决定类别的弱分类器的分类处理通过流水线处理来执行。

公开(公告)号：US09275466B2

公开(公告)日：2016-03-01

申请号：US14508928

申请日：2014-10-07

Applicant: CANON KABUSHIKI KAISHA

Inventor： Tsewei Chen

IPC: G06K9/34 ,  G06T7/00

CPC classification number: G06T7/0081 ,  G06T7/11 ,  G06T7/136

Abstract: An image processing method includes, calculating a partial distance between a pixel of interest in an image and each of reference pixels, sequentially calculating a total distance between the pixel of interest and each of the plurality of the reference pixels based on the partial distance, determining a shortest total distance among the total distances that have been already calculated, in the sequential calculation of the total distance, and categorizing the pixel of interest based on the reference pixel corresponding to the shortest total distance, wherein, if the partial distance between the pixel of interest and a specific one of the reference pixels to be calculated is equal to or greater than the shortest total distance in the sequential calculation of the total distance, the calculation of the total distance between the pixel of interest and the specific one of the reference pixels to be calculated is omitted.

Abstract translation: 一种图像处理方法包括：计算图像中的感兴趣像素与每个参考像素之间的部分距离，基于部分距离顺序地计算感兴趣像素和多个参考像素中的每一个之间的总距离，确定 在总距离的顺序计算中，已经计算出的总距离之间的最短总距离，以及基于对应于最短总距离的参考像素对感兴趣像素进行分类，其中，如果像素之间的部分距离 并且要计算的参考像素中的特定一个等于或大于总距离的顺序计算中的最短总距离，计算感兴趣像素与参考像素中的特定像素之间的总距离 省略要计算的像素。

公开(公告)号：US11755880B2

公开(公告)日：2023-09-12

申请号：US16295384

申请日：2019-03-07

Applicant: CANON KABUSHIKI KAISHA

Inventor： Hongxing Gao ,  Wei Tao ,  Tsewei Chen ,  Dongchao Wen

IPC: G06N3/04 ,  G06N3/08

CPC classification number: G06N3/04 ,  G06N3/08

Abstract: A method and an apparatus for optimizing and applying a multilayer neural network model, and a storage medium are provided. The optimization method includes, dividing out at least one sub-structure from the multilayer neural network model to be optimized, wherein a tail layer of the divided sub-structure is a quantization layer, and transferring operation parameters in layers other than the quantization layer to the quantization layer for each of the divided sub-structures and updating quantization threshold parameters in the quantization layer based on the transferred operation parameters. When a multilayer neural network model optimized based on the optimization method is operated, the necessary processor resources can be reduced.

公开(公告)号：US20210334622A1

公开(公告)日：2021-10-28

申请号：US17230577

申请日：2021-04-14

Applicant: CANON KABUSHIKI KAISHA

Inventor： Wei Tao ,  Tsewei Chen ,  Dongchao Wen ,  Junjie Liu ,  Deyu Wang

IPC: G06N3/04 ,  G06N3/08

Abstract: A method for generating a multilayer neural network including acquiring a multilayer neural network, wherein the multilayer neural network includes at least convolutional layers and quantization layers; generating, for each of the quantization layers in the multilayer neural network, quantization threshold parameters based on a quantization bit parameter and a learnable quantization interval parameter in the quantization layer; and updating the multilayer neural network to obtain a fixed-point neural network based on the generated quantization threshold parameters and operation parameters for each layer in the multilayer neural network.

公开(公告)号：US20210279574A1

公开(公告)日：2021-09-09

申请号：US17189014

申请日：2021-03-01

Applicant: CANON KABUSHIKI KAISHA

Inventor： Junjie Liu ,  Tsewei Chen ,  Dongchao Wen ,  Wei Tao ,  Deyu Wang

IPC: G06N3/08 ,  G06N3/04

Abstract: A method of generating a quantized neural network comprises: determining, based on a floating-point weight in a neural network to be quantized, networks which correspond to the floating-point weights and are used for directly outputting quantized weights, respectively; quantizing, using the determined network, the floating-point weight corresponding to the network to obtain a quantized neural network; updating, based on a loss function value obtained via the quantized neural network, the determined network, the floating-point weight and the quantized weight in the quantized neural network.

公开(公告)号：US20210065011A1

公开(公告)日：2021-03-04

申请号：US17003384

申请日：2020-08-26

Applicant: CANON KABUSHIKI KAISHA

Inventor： Junjie Liu ,  Tsewei Chen ,  Dongchao Wen ,  Wei Tao ,  Deyu Wang

IPC: G06N3/08 ,  G06N3/04 ,  G06N5/04

Abstract: A training and application method, apparatus, system and storage medium of a neural network model is provided. The training method comprises: determining a constraint threshold range according to the number of training iterations and a calculation accuracy of the neural network model, and constraining a gradient of a weight to be within the constraint threshold range, so that when the gradient of a low-accuracy weight is distorted due to a quantization error, the distortion of the gradient is corrected by the constraint of the gradient, thereby making the trained network model achieve the expected performance.

公开(公告)号：US10916016B2

公开(公告)日：2021-02-09

申请号：US16140303

申请日：2018-09-24

Applicant: CANON KABUSHIKI KAISHA

Inventor： Xuyu Zhao ,  Tsewei Chen

IPC: G06T7/194 ,  G06T7/73 ,  G06K9/32 ,  G06K9/00 ,  G06K9/38 ,  G06T7/11

Abstract: Acquiring a current image from an inputted video and a background model which comprises a background image and foreground/background classification information of visual elements; classifying the visual elements in the current image as foreground or background; determining similarity measures between the current image and groups in the background model, wherein visual elements in the current image are the visual elements in the current image which are classified as the foreground, wherein visual elements in the groups in the background model are the visual elements whose classification information is the foreground, and wherein the visual elements in the groups in the background model are the visual elements which neighbour to corresponding portions of the visual elements in the groups in the current image; and identifying whether the visual elements in the current image which are classified as the foreground are falsely classified or not according to the determined similarity measures.

公开(公告)号：US20210019917A1

公开(公告)日：2021-01-21

申请号：US16925801

申请日：2020-07-10

Applicant: CANON KABUSHIKI KAISHA

Inventor： Masahiro Ariizumi ,  Tsewei Chen

IPC: G06T9/00 ,  G06T1/60 ,  G06T7/00 ,  G06T7/174

Abstract: An image processing apparatus comprising, a first processing unit configured to process a first image stored in a first memory and output a first processing result in a first size, a conversion unit configured to, if the first size matches a second size, output the first processing result, and if the first size is different from the second size, convert the first processing result into the second size and output a result of the conversion, and a second processing unit configured to process the first processing result outputted from the conversion unit and a second image of the second size stored in a second memory, and to store a second processing result in the second memory in the second size.

公开(公告)号：US20210004681A1

公开(公告)日：2021-01-07

申请号：US16918206

申请日：2020-07-01

Applicant: CANON KABUSHIKI KAISHA

Inventor： Shunta Tate ,  Tsewei Chen

IPC: G06N3/08 ,  G06N3/04

Abstract: There is provided with a data processing apparatus for detecting an object from an image using a hierarchical neural network. The data processing apparatus has parallel first and second neural networks. An obtaining unit obtains a table which defines different first and second portions. An operation unit performs calculation of the feature data of a third portion based on feature data of the first portion identified using the table and on a weighting parameter between first and second layers of the first neural network, and calculation of feature data of a fourth portion based on feature data of the second portion identified using the table and on a weighting parameter between the first and second layers of the second neural network.