Processing

Please wait...

PATENTSCOPE will be unavailable a few hours for maintenance reason on Tuesday 26.10.2021 at 12:00 PM CEST
Settings

Settings

Goto Application

1. WO2019231105 - METHOD AND APPARATUS FOR LEARNING DEEP LEARNING MODEL FOR ORDINAL CLASSIFICATION PROBLEM BY USING TRIPLET LOSS FUNCTION

Publication Number WO/2019/231105
Publication Date 05.12.2019
International Application No. PCT/KR2019/004452
International Filing Date 12.04.2019
IPC
G06N 3/08 2006.1
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
G06N 3/04 2006.1
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
04Architecture, e.g. interconnection topology
G06K 9/62 2006.1
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 9/62
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
G06N 3/04
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
G06N 3/08
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
3Computer systems based on biological models
02using neural network models
08Learning methods
Applicants
  • 한국과학기술원 KOREA ADVANCED INSTITUTE OF SCIENCE AND TECHNOLOGY [KR]/[KR]
Inventors
  • 양현승 YANG, Hyun-Seung
  • 임우빈 IM, Woobin
  • 홍성은 HONG, Sungeun
  • 윤성의 YOON, Sung-Eui
Agents
  • 특허법인 충현 CHUNG HYUN PATENT & LAW FIRM
Priority Data
10-2018-006270531.05.2018KR
10-2019-004301912.04.2019KR
Publication Language Korean (KO)
Filing Language Korean (KO)
Designated States
Title
(EN) METHOD AND APPARATUS FOR LEARNING DEEP LEARNING MODEL FOR ORDINAL CLASSIFICATION PROBLEM BY USING TRIPLET LOSS FUNCTION
(FR) PROCÉDÉ ET APPAREIL POUR APPRENDRE UN MODÈLE D'APPRENTISSAGE PROFOND POUR UN PROBLÈME DE CLASSIFICATION ORDINAL À L'AIDE D'UNE FONCTION DE PERTE DE TRIPLET
(KO) 트리플릿 기반의 손실함수를 활용한 순서가 있는 분류문제를 위한 딥러닝 모델 학습 방법 및 장치
Abstract
(EN)
The present invention relates to image processing using machine learning, and a method for learning a deep learning model for an ordinal classification problem makes a learning object into an input; forms convolutional neural networks (CNNs) including a branch point and two end points, which are separated from a branch thereof so as to cause classification loss and triplet loss, calculates classification loss for end-to-end learning, calculates the triplet loss such that a network can learn ordinal characteristics, and updates the network for a final loss value by performing relative triplet sampling on the basis of the calculated classification loss and triplet loss, thereby enabling effective learning and loss control.
(FR)
La présente invention concerne un traitement d'image à l'aide d'un apprentissage automatique, et un procédé d'apprentissage d'un modèle d'apprentissage profond pour un problème de classification ordinal comprenant les étapes consistant à convertir un objet d'apprentissage en entrée ; à former des réseaux neuronaux convolutionnels (CNN) comprenant un point de ramification et deux points d'extrémité qui sont séparés d'une branche de ces derniers de façon à provoquer une perte de classification et une perte de triplet, à calculer une perte de classification pour un apprentissage de bout en bout, à calculer la perte de triplet de telle sorte qu'un réseau peut apprendre des caractéristiques ordinales, et à mettre à jour le réseau pour une valeur de perte finale en effectuant un échantillonnage de triplet relatif sur la base de la perte de classification et de la perte de triplet calculées, ce qui permet un apprentissage efficace et une commande de perte.
(KO)
본 발명은 기계 학습을 이용한 영상 처리에 관한 기술로, 순서가 있는 분류 문제를 위한 딥러닝 모델을 학습하는 방법은, 학습 대상을 입력으로 하고 분기점과 그 분기에서 나누어져 분류 손실(classification loss)과 트리플릿 손실(triplet loss)을 발생시키는 두 개의 종단점으로 구성된 CNN(Convolutional Neural Networks)을 형성하고, 종단간(end-to-end) 학습을 위한 분류 손실을 산출하고, 네트워크가 순서 특성을 학습할 수 있도록 트리플릿 손실을 산출하며, 산출된 분류 손실 및 트리플릿 손실에 기반하되 상관 트리플릿 샘플링(relative triplet sampling)을 수행함으로써 최종 손실값에 대해 네트워크를 갱신함으로써, 효과적인 학습과 손실 제어가 가능하다.
Related patent documents
Latest bibliographic data on file with the International Bureau