Class estimation apparatus, non-transitory computer readable medium, and class estimation method

• US 10,303,985 B2
• Filed: 10/13/2016
• Issued: 05/28/2019
• Est. Priority Date: 05/20/2016
• Status: Active Grant

• Current Assignee: Fujifilm Business Innovation Corp. (Fujifilm Holdings Corporation)
• Original Assignee: Fuji Xerox Company Limited (Xerox Holdings Corp.)
• Inventors: Ozaki, Ryota, Tsuboshita, Yukihiro, Kato, Noriji
• Primary Examiner(s): Bernardi, Brenda C
• Application Number:
  US15/292,528
• Publication Number:
  US 20170337450A1
• Time in Patent Office: 957 Days
• CPC Class Codes: G06F 18/2415 based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
• CPC Class Codes: G06F 18/24323 Tree-organised classifiers
• CPC Class Codes: G06F 18/285 Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system

Abstract

There is provided a class estimation apparatus. A data reception unit receives an input of a feature amount of data. An identification device reception unit receives an input of an identification device having a tree structure. A table storage unit stores a probability table and a feature amount table in a leaf node of the tree structure. A probability table correction unit corrects the probability table. A class estimation unit estimates a class of the data.

Claims

1. A non-transitory computer readable medium storing a program causing a computer to execute:

a data reception step of receiving an input of image data including feature amounts;

an identification device reception step of receiving an input of an identification device having a tree structure;

a table storage step of storing a probability table and a feature amount table in a leaf node of the tree structure;

a probability table correcting step of correcting the probability table;

wherein,

when input image data for which a belonging class is unknown is input to the computer, the computer searches a leaf node to which the input image data belongs; and

corrects the probability table of the leaf node to which the input image data belongs,

wherein, when a class in which likelihood calculated from a value of the probability table is greatest and a class in which a distance calculated from a feature amount of the input image data and a value of the feature amount table is shortest are same, the computer sets the number of data of the classes except for the class in which the likelihood is greatest and the distance is shortest to zero;

a class estimation step of estimating a class of the input image data using the probability table corrected in the probability table correcting step; and

an output step of outputting an estimation result based on the class estimating step.

3. The non-transitory computer readable medium according to claim 1, wherein the computer further executes:

storing an importance degree in the leaf node of the tree structure.

2. A class estimation method executed by a computer, comprising:

receiving an input of image data including feature amounts;

receiving an input of an identification device having a tree structure;

storing a probability table and a feature amount table in a leaf node of the tree structure;

correcting the probability table, wherein,

when input image data for which a belonging class is unknown is input to the computer, the computer searches a leaf node to which the input image data belongs; and

corrects the probability table of the leaf node to which the input image data belongs, wherein, when a class in which likelihood calculated from a value of the probability table is greatest and a class in which a distance calculated from a feature amount of the input image data and a value of the feature amount table is shortest are same, the computer sets the number of data of the classes except for the class in which the likelihood is greatest and the distance is shortest to zero;

estimating a class of the input image data using the probability table corrected in the probability table correcting step; and

outputting an estimation result based on the estimating step.

4. The class estimation method according to claim 2, wherein the computer further executes:

storing an importance degree in the leaf node of the tree structure.