Fuzzy associative system for multimedia object description

• US 7,460,715 B2
• Filed: 08/27/2003
• Issued: 12/02/2008
• Est. Priority Date: 08/30/2002
• Status: Expired due to Fees

• Current Assignee: Orange S.A.
• Original Assignee: Orange S.A.
• Inventors: Lancieri, Luigi
• Primary Examiner(s): Bella; Matthew C
• Assistant Examiner(s): PERUNGAVOOR, SATHYANARAYA V
• Application Number:
  US10/526,090
• Publication Number:
  US 20050249418A1
• Time in Patent Office: 1,924 Days
• Field of Search: 382/228, 382/224
• US Class Current: 382/224
• CPC Class Codes: G06F 16/48 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 18/00 Pattern recognition

Abstract

The invention concerns a method for automatic description of an unknown multimedia object, which consists in associating the unknown object with several types of reference multimedia objects each time based on a membership probability particular to each type concerned (G), the method including a step which consists in measuring at least one physical characteristic on the unknown object (F) and in comparing it with measurements of characteristics representing best the reference types, the method further comprising the step (H) which consists in using for each type at least one probabilistic relationship providing a membership probability to the type concerned based on the result of the comparison of the characteristics of the unknown object and of the type, the method further comprising the step which consists in using the probabilities of membership to the various types thus obtained in combination with a series of affinity relationships between types, so as to select memberships which both constitute the majority in terms of probability and which designate one another by their affinity relationship, and so as to exclude memberships with less affinity with the selected types.

Claims

1. A method for automatically describing an unknown multimedia object in which the unknown object is associated with a plurality of reference multimedia objects depending on a probability of membership to each considered type of said plural reference multimedia objects such that multimedia objects having memberships in each considered type of said plural reference multimedia objects and a lower affinity to elected multimedia objects having memberships in each type of said plural reference multimedia objects are excluded, the method comprising the steps of:

measuring at least one physical characteristic of the unknown multimedia object and comparing the at least one physical characteristic of the unknown multimedia object with measurements of characteristics which best represent each type of said plural reference multimedia objects;

using at least one probabilistic relation for each type of said plural reference multimedia objects to obtain a probability that the unknown multimedia object is a member of the type of said plural reference multimedia objects considered based on a result of the comparison of the at least one measured physical characteristics of the unknown object to the measurements of the characteristics which best represent each type of said plural reference multimedia objects;

electing the multimedia objects having memberships in each type of said plural reference multimedia objects which form a majority of probabilities associated with each type of said plural reference multimedia objects considered and which are co-designated by an associated affinity relation using obtained probabilities that the unknown multimedia objects are members of different types of said plural reference multimedia objects in combination with a series of affinity relations between each type of said plural reference multimedia objects such that the multimedia objects having memberships in each considered type of said plural reference multimedia objects and the lower affinity to the elected multimedia objects having memberships in each type of said plural reference multimedia objects are excluded.

2. The method according to claim 1, further comprising a preliminary step of:

defining reference characteristics of a type of said plural reference multimedia objects starting from a group of multimedia objects presumed to represent each type of said plural reference multimedia objects, said step of defining the reference characteristics comprising;

measuring a physical characteristic of the group of multimedia objects; and

obtaining at least one reference value for the measured characteristic, said at least one reference value being used to define the probability of membership relation to the type of said plural reference multimedia objects as a value to which the measurement of the at least one physical characteristic of the unknown object is compared to determine the probability of membership to the type of said plural reference multimedia objects.

3. The method according to claim 2, further comprising the step of:

performing an automatic search in an information system with an Internet downloader and an Internet search engine to provide the group of multimedia objects.

4. The method according to claim 1, further comprising a preliminary step of:

measuring a frequency of simultaneous occurrence of types of said plural reference multimedia object in a set of descriptive objects to determine existence of an affinity between at least two types of said plural reference multimedia objects when the types of said plural reference multimedia objects have a specific simultaneous rate of occurrence.

5. The method according to claim 1, further comprising the step of:

recognizing a shape of the unknown multimedia object;

wherein at least one reference shape of the unknown multimedia object to be recognized comprises a physical characteristic belonging to a definition of one of the types of said plural reference multimedia objects.

6. The method according to claim 5, wherein the shape recognizing step includes a similarity measurement between a determined shape of the unknown multimedia object and the at least one reference shape, and utilization of a predefined relation which provides a probability of membership of the unknown multimedia object to the type of said plural reference multimedia objects as a function of the shape similarity measurement made between the determined shape of the unknown multimedia object and the at least one reference shape.

7. The method according to claim 1, wherein the at least one type of physical characteristic of the unknown multimedia object includes a plurality of reference characteristics;

wherein at least two measurements are performed on the unknown multimedia object to generate a proximity measurement with each of said plural reference characteristics;

the probability of membership in the type of reference characteristics is performed using at least two relations, each providing a probability of membership to the type of reference characteristics as a function of a proximity to a different characteristic;

wherein the at least two probability relations are used to set up a resultant global probability of membership of the object to the type of reference characteristics considered.

8. The method according to claim 7, wherein the at least two relations which each provide the probability of membership to the type of reference characteristics are used in accordance with combinational fuzzy logic to provide the resultant global probability of membership of the unknown multimedia object to the type of reference characteristics considered.

9. The method according to claim 1, further comprising the step of:

implementing fuzzy logic comprising a mechanism that provides a single probability of membership to a reference type of characteristic starting from a combination of similarity probabilities with different characteristics to the reference type of characteristic.

10. A device for automatically providing a description of an unknown multimedia object in which the unknown object is associated with a plurality of reference multimedia objects depending on a probability of membership to a considered type of said plural reference multimedia objects the device comprising:

means for measuring at least one physical characteristic of the unknown object and comparing the at least one physical characteristic of the unknown object with measurements of characteristics which best represent each type of said plural reference multimedia objects;

means for using at least one probabilistic relation for each type of said plural reference multimedia objects to provide a probability that the unknown multimedia object is a member of the type of said plural reference multimedia objects considered based on a result of the comparison of the at least one measure physical characteristics of the unknown object to the measurements of the characteristics which best represent each type of said plural reference multimedia objects;

means for electing multimedia objects having memberships in each type of said plural reference multimedia objects which form a majority of probabilities associated with each type of said plural reference multimedia objects considered and which are co-designated by an associated affinity relation using obtained probabilities that the unknown multimedia objects are members of different types of said plural reference multimedia objects in combination with a series of affinity relations between each type of said plural reference multimedia objects such that multimedia objects having memberships in the type of said plural reference multimedia objects having a lower affinity to the elected multimedia objects in the type of plural reference multimedia objects are excluded.

11. The device according to claim 10, further comprising:

processing means for utilizing a plurality of groups of reference multimedia objects, each of said plural groups providing a best representation of a corresponding type of said plural reference multimedia objects, said processing means being configured to measure at least one physical characteristic for an entire group considered, and for deriving a reference measurement for the at least one physical characteristic for the entire group considered from the at least one physical characteristic, the derived reference measurement being utilized to define a relationship providing a probability of membership to the type of said plural reference multimedia objects considered as a measurement to which the device compares a measurement of an unknown object to determine the probability of membership to the type of said plural reference multimedia objects considered.

12. The device according to claim 10, further comprising:

means for preliminarily measuring a frequency of occurrence of types of said plural reference multimedia object in a set of descriptive objects, and for determining an existence of an affinity between at least two types of said plural reference multimedia objects when types of said plural reference multimedia objects have a specific simultaneous rate of occurrence.

13. The device according to claim 11, further comprising:

means for preliminarily measuring a frequency of occurrence of types of said plural reference multimedia object in a set of descriptive objects, and for determining an existence of an affinity between at least two types of said plural reference multimedia objects when types of said plural reference multimedia objects have a specific simultaneous rate of occurrence.