System and method for recognizing characters in multimedia content
A system and method for recognizing characters embedded in multimedia content are provided. The method includes extracting at least one image of at least one character from a received multimedia content item; identifying a natural language character corresponding to the at least one image of the at least one character, wherein the identification is performed by a deep content classification (DCC) system; and storing the identified natural language character in a data warehouse.
1. A method for recognizing characters embedded in multimedia content, comprising:
extracting at least one image of at least one character from a received multimedia content item;
identifying a natural language character corresponding to the at least one image of the at least one character, wherein the identification is performed by a deep content classification (DCC) system;
comparing a first signature of a first image of at least one first natural language character to at least one second signature of at least one second multimedia content item within a concept structure, a matching concept structure being identified if matching between the first signature of the first image and the at least one second signature of the at least one second multimedia content item within the concept structure is above a first threshold;
comparing the first signature of the first image to a plurality of signatures of a plurality of natural language words associated with the matching concept structure until a match is found above a second threshold yielding a matching natural language word of the plurality of natural words;
wherein the second threshold exceeds the first threshold; and
storing the identified natural language character in a data warehouse.
2. The method of claim 1 , further comprising:
comparing a sequence of the natural language character to a library of natural language words;
detecting a natural language word in the library of natural language words that matches the sequence of natural language characters; and
storing the detected natural language word in the data warehouse.
3. The method of claim 1 , wherein the multimedia content item is received from a user device.
4. The method of claim 3 , further comprising: sending the natural language character to the user device.
5. The method of claim 1 , wherein the at least one multimedia content item is any one of: an image, a graphic, a video stream, a video clip, a video frame, and a photograph.
6. The method of claim 1 , further comprising:
generating at least one signature respective of the at least one character; and
querying the DCC system using the at least on generated signature to identify the natural language character corresponding to the image of the at least one character.
7. The method of claim 6 , further comprising: identifying a concept structure corresponding to the image based on the at least one signature, wherein the identified natural language character is associated with the identified concept structure.
8. The method of claim 1 , wherein each image of the extracted at least one image is extracted based on an extraction order.
9. The method of claim 1 , wherein the extraction order is at least any of: left to right, right to left, top to bottom, bottom to top, paragraphs before sentences, sentences before phrases, and phrases before words.
10. A non-transitory computer readable medium having stored thereon instructions for causing one or more processing units to execute the method according to claim 1 .
11. A system for recognizing characters embedded in multimedia content, comprising:
an interface to a network for receiving a multimedia content item;
a processor;
a memory connected to the processor,
wherein the memory contains instructions that, when executed by the processor, configure the system to:
extract at least one image of at least one character from the received multimedia content item;
identify a natural language character corresponding to the at least one image of the at least one character, wherein the identification is performed by a deep content classification (DCC) system;
compare a first signature of a first image of at least one first natural language character to at least one second signature of at least one second multimedia content item within a concept structure, a matching concept structure being identified if matching between the first signature of the first image and the at least one second signature of the at least one second multimedia content item within the concept structure is above a first threshold;
compare the first signature of the first image to a plurality of signatures of a plurality of natural language words associated with the matching concept structure until a match is found above a second threshold yielding a matching natural language word of the plurality of natural words; wherein the second threshold exceeds the first threshold;
and store the identified natural language character in a data warehouse.
12. The system of claim 11 , wherein the system is further configured to: compare a sequence of the at least a natural language character to a library of natural language words; detect a natural language word in the library of natural language words that matches the sequence of natural language characters; and store the detected natural language word in the data warehouse.
13. The system of claim 11 , wherein the multimedia content item is received from a user device.
14. The system of claim 13 , wherein the system is further configured to: send the natural language character to the user device.
15. The system of claim 11 , wherein the at least one multimedia content item is any one of: an image, a graphic, a video stream, a video clip, a video frame, and a photograph.
16. The system of claim 11 , wherein the system is further configured to: generate at least one signature respective of the at least one character; and querying the DCC system using the at least on generated signature to identify the natural language character corresponding to the image of the at least one character.
17. The system of claim 16 , wherein the system is further configured to: identify a concept structure corresponding to the image based on the at least one signature, wherein the identified natural language character is associated with the identified concept structure.
18. The system of claim 11 , wherein each image of the extracted at least one image is extracted based on an extraction order.
19. The system of claim 11 , wherein the extraction order is at least any one of: left to right, right to left, top to bottom, bottom to top, paragraphs before sentences, sentences before phrases, and phrases before words.
20. The method according to claim 1 wherein the first signature represents a response of leaky integrate-to-threshold unit nodes to the first image.
21. The method according to claim 1 comprising generating the first signature by a plurality of mutually independent computational cores.
22. The method according to claim 1 wherein the concept structure comprises signatures of objects of different type that share a property.
23. The method according to claim 1 wherein the concept structure comprises signature reduced cluster.
24. The method according to claim 1 wherein the concept structure comprises signatures of different types of motor vehicles.