Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020141587 - INFORMATION PROCESSING DEVICE, INFORMATION PROCESSING METHOD, AND PROGRAM

Publication Number WO/2020/141587
Publication Date 09.07.2020
International Application No. PCT/JP2019/048321
International Filing Date 10.12.2019
IPC
G06N 20/00 2019.1
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
20Machine learning
CPC
G06K 9/6217
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
6217Design or setup of recognition systems and techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
G06K 9/6277
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
6277based on a parametric (probabilistic) model, e.g. based on Neyman-Pearson lemma, likelihood ratio, Receiver Operating Characteristic [ROC] curve plotting a False Acceptance Rate [FAR] versus a False Reject Rate [FRR]
G06K 9/628
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
6279relating to the number of classes
628Multiple classes
G06N 20/00
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
20Machine learning
G06N 5/04
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
5Computer systems using knowledge-based models
04Inference methods or devices
Applicants
  • パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカ PANASONIC INTELLECTUAL PROPERTY CORPORATION OF AMERICA [US]/[US]
Inventors
  • 藤村 亮太 FUJIMURA, Ryota
  • 中田 洋平 NAKATA, Yohei
  • 築澤 宗太郎 TSUKIZAWA, Sotaro
Agents
  • 新居 広守 NII, Hiromori
  • 寺谷 英作 TERATANI, Eisaku
  • 道坂 伸一 MICHISAKA, Shinichi
Priority Data
2019-10787810.06.2019JP
62/787,57602.01.2019US
Publication Language Japanese (JA)
Filing Language Japanese (JA)
Designated States
Title
(EN) INFORMATION PROCESSING DEVICE, INFORMATION PROCESSING METHOD, AND PROGRAM
(FR) DISPOSITIF ET PROCÉDÉ DE TRAITEMENT D'INFORMATIONS, AINSI QUE PROGRAMME
(JA) 情報処理装置、情報処理方法及びプログラム
Abstract
(EN)
This information processing device (100a) is an information processing device comprising a processor, wherein: the processor acquires a first classification threshold for classifying data under at least one of a plurality of classes, and outputs a classification result in which the data has been classified under at least one of the plurality of classes on the basis of the first classification threshold and an output of a trained classification model; the first classification threshold is obtained by subjecting a second classification threshold to a second transform which is an inverse transform of a first transform; the first transform is a transform from the output of the trained classification model to a classification probability value of a plurality of singleton classes constituting the plurality of classes; and the second classification threshold is set on the basis of the classification probability value of the plurality of singleton classes.
(FR)
La présente invention concerne un dispositif de traitement d'informations (100a) qui est un dispositif de traitement d'informations comprenant un processeur, le processeur acquérant un premier seuil de classification pour classifier des données sous au moins une classe parmi une pluralité de classes, et délivrant un résultat de classification selon lequel les données ont été classifiées dans au moins l'une de la pluralité de classes sur la base du premier seuil de classification et d'une sortie d'un modèle de classification entraîné ; le premier seuil de classification étant obtenu en soumettant un second seuil de classification à une seconde transformée qui est une transformée inverse d'une première transformée ; la première transformée étant une transformée de la sortie du modèle de classification entraîné à une valeur de probabilité de classification d'une pluralité de classes de singleton constituant la pluralité de classes ; et le second seuil de classification étant défini sur la base de la valeur de probabilité de classification de la pluralité de classes de singleton.
(JA)
情報処理装置(100a)は、プロセッサを備える情報処理装置であって、プロセッサは、データを複数のクラスの少なくとも1つに分類するための第1分類閾値を取得し、訓練済みの分類モデルの出力と第1分類閾値とに基づいてデータを複数のクラスの少なくとも1つに分類した分類結果を出力し、第1分類閾値は、第2分類閾値の、第1変換に対する逆変換である第2変換により得られ、第1変換は、訓練済みの分類モデルの出力から複数のクラスを構成する複数の単体クラスの分類確率値への変換であり、第2分類閾値は、複数の単体クラスの分類確率値に基づいて設定される。
Also published as
JP2020563863
Latest bibliographic data on file with the International Bureau