Information processing apparatus, information processing method, and storage medium

Patent metadata

Field	Value
Publication number	US-2018181885-A1
Application number	US-201715845922-A
Country	US
Kind code	A1
Filing date	Dec 18, 2017
Priority date	Dec 22, 2016
Publication date	Jun 28, 2018
Grant date	—

Abstract

An information processing apparatus includes an obtaining unit configured to obtain information of a plurality of labels applied to the learning data by a plurality of users, information regarding reliability of each applied label itself, and information regarding reliability of a user who applies the relevant label, wherein the information of the label is information regarding a result to be recognized in a case where the predetermined recognition is performed on the learning data and a determination unit configured to determine a label to the learning data from among the plurality of labels based on the reliability of the label itself and the reliability of the user who applies the relevant label.

First claim

What is claimed is:

1 . An information processing apparatus which processes learning data used in learning of a dictionary for performing predetermined recognition, the information processing apparatus comprising:
an obtaining unit configured to obtain information of a plurality of labels applied to the learning data by a plurality of users, information regarding reliability of each applied label itself, and information regarding reliability of a user who applies the relevant label, wherein the information of the label is information regarding a result to be recognized in a case where the predetermined recognition is performed on the learning data; and a determination unit configured to determine a label to the learning data from among the plurality of labels based on the reliability of the label itself and the reliability of the user who applies the relevant label.

2 . The information processing apparatus according to claim 1 , further comprising an evaluation unit configured to derive an evaluation value for each of the plurality of labels applied to the learning data,
wherein the determination unit determines a label to the learning data based on the evaluation value.

3 . The information processing apparatus according to claim 1 , wherein the determination unit determines a label to the learning data further based on a likelihood of a result that a classifier for identifying a label of learning data applies to the learning data.

4 . The information processing apparatus according to claim 1 , further comprising a display control unit configured to display the plurality of labels applied to the learning data on a display apparatus.

5 . The information processing apparatus according to claim 4 , further comprising a label comparison unit configured to compare a first label and a second label in the plurality of labels applied to the learning data,
wherein the display control unit displays a result compared by the label comparison unit on the display apparatus.

6 . The information processing apparatus according to claim 5 , wherein
the label comparison unit calculates a similarity degree or a coincidence degree between the plurality of labels, and the display control unit displays a derived similarity degree or a derived coincidence degree.

7 . The information processing apparatus according to claim 4 , wherein the display control unit has a function of displaying the label by filtering or sorting based on the reliability of the label itself.

8 . The information processing apparatus according to claim 4 , wherein the display control unit has a function of displaying the label by filtering or sorting based on the reliability of the user who applies the label.

9 . The information processing apparatus according to claim 4 , wherein the display control unit displays another labeled learning data piece related to learning data which a user labels on the display apparatus.

10 . The information processing apparatus according to claim 1 , further comprising a certainty degree derivation unit configured to calculate a certainty degree representing a likelihood of the determined label based on the reliability of the label itself.

11 . The information processing apparatus according to claim 10 , further comprising a learning unit configured to learn a prediction model based on the determined label; and
a recognition unit configured to perform recognition using the prediction model.

12 . The information processing apparatus according to claim 11 , wherein the certainty degree derivation unit derives a certainty degree based on a result recognized by the recognition unit.

13 . The information processing apparatus according to claim 10 , further comprising a data extraction unit configured to extract a data group regarding which a request for labeling is made from among a learning data group based on the certainty degree.

14 . The information processing apparatus according to claim 13 , wherein the data extraction unit extracts data of which a certainty degree is low.

15 . The information processing apparatus according to claim 14 , wherein the data extraction unit extracts the data group so that the data group includes a predetermined number or more pieces of learning data which are already labeled.

16 . The information processing apparatus according to claim 1 , further comprising a label evaluation unit configured to evaluate labeling applied to the learning data based on a calculated result of a similarity degree between the label applied to the learning data and the label determined by the determination unit.

17 . The information processing apparatus according to claim 16 , wherein a label evaluation unit derives reliability of a person or an algorithm which performs labeling based on the label evaluated by the label evaluation unit.

18 . A method of information processing for processing learning data used in learning of a dictionary for performing predetermined recognition, the method comprising:
obtaining information of a plurality of labels applied to the learning data by a plurality of users, information regarding reliability of each applied label itself, and information regarding reliability of a user who applies the relevant label, wherein the information of the label is information regarding a result to be recognized in a case where the predetermined recognition is performed on the learning data; and determining a label to the learning data from among the plurality of labels based on the reliability of the label itself and the reliability of the user who applies the relevant label.

19 . A storage medium storing a program for causing a computer to execute each step in a method of information processing for processing learning data used in learning of a dictionary for performing predetermined recognition, the method comprising:
obtaining information of a plurality of labels applied to the learning data by a plurality of users, information regarding reliability of each applied label itself, and information regarding reliability of a user who applies the relevant label, wherein the information of the label is information regarding a result to be recognized in a case where the predetermined recognition is performed on the learning data; and determining a label to the learning data from among the plurality of labels based on the reliability of the label itself and the reliability of the user who applies the relevant label.

Assignees

• Canon Kk

Inventors

• Higo Tomoaki
• Inaba Masaki
• Yamada Takayuki
• Suzuki Masahiro

Classifications

• G06F40/242

  Dictionaries · CPC title

• G06F40/279

  Recognition of textual entities · CPC title

• G06F16/9017

  using directory or table look-up (use of a directory or look-up table in file systems G06F16/13) · CPC title

• G06N20/20

  Ensemble learning · CPC title

• G06F17/2735

  Physics · mapped topic