Information processing apparatus, information processing method, and non-transitory computer readable storage medium

• US 9,824,301 B2
• Filed: 04/27/2016
• Issued: 11/21/2017
• Est. Priority Date: 04/30/2015
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Chen, Tsewei
• Primary Examiner(s): Chawan, Sheela C
• Application Number:
  US15/139,603
• Publication Number:
  US 20160321521A1
• Time in Patent Office: 573 Days
• Field of Search: 382224, 382225, 382227, 382303, 382304
• CPC Class Codes: G06F 18/2148 characterised by the process organisation or structure, e.g. boosting cascade
• CPC Class Codes: G06F 18/24323 Tree-organised classifiers
• CPC Class Codes: G06F 18/2451 linear, e.g. hyperplane
• CPC Class Codes: G06V 10/764 using classification, e.g. of video objects
• CPC Class Codes: G06V 10/7747 Organisation of the process, e.g. bagging or boosting
• CPC Class Codes: G06V 40/161 Detection; Localisation; Normalisation

Abstract

In an information processing apparatus that includes sequences of weak classifiers which are logically cascade-connected in each sequence and the sequences respectively correspond to categories of an object and in which the weak classifiers are grouped into at least a first group and a second group in the order of connection, classification processing by weak classifiers belonging to the first group of respective categories is performed by pipeline processing. Based on the processing results of the weak classifiers belonging to the first group of the respective categories, categories in which classification processing by weak classifiers belonging to the second group is to be performed are decided out of the categories. The classification processing by the weak classifiers respectively corresponding to the decided categories and belonging to the second group is performed by pipeline processing.

Claims

1. An information processing apparatus comprising:

at least one memory which stores instructions; and

one or more processors which execute the instructions to cause the information processing apparatus to;

implement a plurality of sequences of weak classifiers which are logically cascade-connected in each sequence, wherein the sequences respectively correspond to a plurality of categories of an object and the plurality of weak classifiers are grouped into at least a first group and a second group in the order of connection;

control to perform, by pipeline processing, classification processing by weak classifiers belonging to the first group in the sequences respectively corresponding to the plurality of categories;

decide, based on results of classification processing by the weak classifiers belonging to the first group in the sequences respectively corresponding to the plurality of categories, categories for which classification processing are to be performed by weak classifiers belonging to the second group out of the plurality of categories; and

control to perform, by pipeline processing, classification processing by the weak classifiers belonging to the second group in sequences respectively corresponding to the decided categories.

2. The apparatus according to claim 1, wherein the one or more processors further cause the information processing apparatus to obtain evaluation values of the results of classification processing by the weak classifiers belonging to the first group in sequences respectively corresponding to the plurality of categories and decide, based on the evaluation values, a category for which classification processing are to be performed by the weak classifiers belonging to the second group out of the plurality of categories.

3. The apparatus according to claim 2, wherein the one or more processors further cause the information processing apparatus to decide categories in which the respective evaluation values exceed a threshold as the categories for which classification processing are to be performed by the respective weak classifiers belonging to the second group.

4. The apparatus according to claim 2, wherein the first group includes a plurality of weak classifiers in a sequence corresponding to the same category, and

the one or more processors further causes the information processing apparatus to obtain evaluation values of results of classification processing by the respective weak classifiers belonging to the first group in a sequence corresponding to a given category and decide the given category as the category for which classification processing are to be performed by the weak classifiers belonging to the second group if any of the evaluation values of the results of classification processing by the respective weak classifiers exceeds a threshold.

5. The apparatus according to claim 2, wherein the first group includes a plurality of weak classifiers for each category, and

wherein the one or more processors further cause the information processing apparatus to obtain evaluation values of results of classification processing by the respective weak classifiers belonging to the first group in a sequence corresponding to a given category and decide the given category as the category for which classification processing are to be performed by the weak classifiers belonging to the second group if a total value of the evaluation values of the results of the classification processing by the respective weak classifiers exceeds a threshold.

6. The apparatus according to claim 1, wherein the one or more processors further cause the information processing apparatus to control to first perform, by pipeline processing, classification processing by the weak classifiers belonging to the first group in a sequence of a category preset as a category having a high possibility that classification processing by the weak classifiers belonging to the second group will be performed.

7. The apparatus according to claim 1, wherein the one or more processors further cause the information processing apparatus to control to start classification processing by the weak classifiers belonging to the second group prior to completion of classification processing by all of the weak classifiers belonging to the first group.

8. An information processing method performed by an information processing apparatus that includes a plurality of sequences of weak classifiers which are logically cascade-connected in each sequence and the sequences respectively correspond to a plurality of categories of an object and in which the plurality of weak classifiers are grouped into at least a first group and a second group in the order of connection, comprising:

performing, by pipeline processing, classification processing by weak classifiers belonging to the first group in the sequences respectively corresponding to the plurality of categories;

deciding, based on results of classification processing by the weak classifiers belonging to the first group in the sequences respectively corresponding to the plurality of categories, categories for which classification processing are to be performed by weak classifiers belonging to the second group out of the plurality of categories; and

performing, by pipeline processing, classification processing by the weak classifiers belonging to the second group in sequences respectively corresponding to the decided categories.

9. A non-transitory computer-readable storage medium storing a computer program for causing a computer to:

implement a plurality of sequences of weak classifiers which are logically cascade-connected in each sequence, wherein the sequences respectively correspond to a plurality of categories of an object and the plurality of weak classifiers are grouped into at least a first group and a second group in the order of connection;

control to perform, by pipeline processing, classification processing by weak classifiers belonging to the first group in the sequences respectively corresponding to the plurality of categories;

decide, based on results of classification processing by the weak classifiers belonging to the first group in the sequences respectively corresponding to the plurality of categories, categories for which classification processing are to be performed by weak classifiers belonging to the second group out of the plurality of categories; and

control to perform, by pipeline processing, classification processing by the weak classifiers belonging to the second group in sequences respectively corresponding to the decided categories.