Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020108474 - PICTURE CLASSIFICATION METHOD, CLASSIFICATION IDENTIFICATION MODEL GENERATION METHOD AND APPARATUS, DEVICE, AND MEDIUM

Publication Number WO/2020/108474
Publication Date 04.06.2020
International Application No. PCT/CN2019/120903
International Filing Date 26.11.2019
IPC
G06K 9/62 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
62Methods or arrangements for recognition using electronic means
CPC
G06K 9/6268
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
62Methods or arrangements for recognition using electronic means
6267Classification techniques
6268relating to the classification paradigm, e.g. parametric or non-parametric approaches
G06N 3/0454
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
04Architectures, e.g. interconnection topology
0454using a combination of multiple neural nets
G06N 3/084
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
084Back-propagation
Applicants
  • 广州市百果园信息技术有限公司 GUANGZHOU BAIGUOYUAN INFORMATION TECHNOLOGY CO., LTD. [CN]/[CN]
Inventors
  • 潘跃 PAN, Yue
  • 刘振强 LIU, Zhenqiang
  • 梁柱锦 LIANG, Zhujin
Agents
  • 北京品源专利代理有限公司 BEYOND ATTORNEYS AT LAW
Priority Data
201811457125.130.11.2018CN
Publication Language Chinese (ZH)
Filing Language Chinese (ZH)
Designated States
Title
(EN) PICTURE CLASSIFICATION METHOD, CLASSIFICATION IDENTIFICATION MODEL GENERATION METHOD AND APPARATUS, DEVICE, AND MEDIUM
(FR) PROCÉDÉ DE CLASSIFICATION D'IMAGES, PROCÉDÉ ET APPAREIL DE GÉNÉRATION DE MODÈLE D'IDENTIFICATION DE CLASSIFICATION, DISPOSITIF ET SUPPORT
(ZH) 图片分类、分类识别模型的生成方法、装置、设备及介质
Abstract
(EN)
A picture classification method, a classification identification model generation method and apparatus, a device, and a medium. The picture classification method comprises: obtaining a picture set to be classified, said picture set comprising at least two pictures (110); inputting said picture set into a pre-trained current level of classification identification model to obtain the classification score of each picture (120); determining whether the classification score of each picture satisfies a preset condition (130); if yes, determining, according to the classification score of each picture, the classification identification result of each picture (140); if not, continuing inputting each picture into a pre-trained next level of classification identification model until the classification score of each picture is obtained (150), each level of classification identification model being generated on the basis of the training of a neural network.
(FR)
L'invention concerne un procédé de classification d'image, un procédé et un appareil de génération de modèle d'identification de classification, un dispositif et un support. Le procédé de classification d'images comprend les étapes consistant à : obtenir un ensemble d'images à classifier, ledit ensemble d'images comprenant au moins deux images (110) ; entrer ledit ensemble d'images dans un niveau pré-entraîné actuel de modèle d'identification de classification afin d'obtenir le score de classification de chaque image (120) ; déterminer si le score de classification de chaque image satisfait une condition prédéfinie (130) ; si tel est le cas, déterminer, en fonction du score de classification de chaque image, le résultat d'identification de classification de chaque image (140) ; dans le cas contraire, continuer à entrer chaque image dans un niveau pré-entraîné suivant de modèle d'identification de classification jusqu'à ce que le score de classification de chaque image soit obtenu (150), chaque niveau de modèle d'identification de classification étant généré sur la base de l'entraînement d'un réseau neuronal.
(ZH)
一种图片分类、分类识别模型的生成方法、装置、设备及介质。该图片分类方法包括:获取待分类的图片集,图片集包括至少两张图片(110);将图片集输入至预先训练的当前级分类识别模型中,得到每张图片的分类得分(120);判断所述每张图片的分类得分满是否足预设条件(130),如果是,则根据所述每张图片的分类得分确定所述每张图片的分类识别结果(140),如果否,则将所述每张图片继续输入至预先训练的下一级分类识别模型中,直至得到所述每张图片的分类得分(150);其中,每级分类识别模型基于神经网络训练生成。
Also published as
Latest bibliographic data on file with the International Bureau