System and method for speech to speech translation using cores of a natural liquid architecture system

• US 9,940,326 B2
• Filed: 10/10/2016
• Issued: 04/10/2018
• Est. Priority Date: 10/26/2005
• Status: Active Grant

• Current Assignee: Cortica Ltd.
• Original Assignee: Cortica Ltd.
• Inventors: Raichelgauz, Igal, Odinaev, Karina, Zeevi, Yehoshua Y
• Primary Examiner(s): Woo, Isaac M
• Application Number:
  US15/289,696
• Publication Number:
  US 20170024379A1
• Time in Patent Office: 547 Days
• Field of Search: 707600-899
• CPC Class Codes: G06F 16/41 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/685 using automatically derived transcript of audio data, e.g. lyrics speech recognition G10L15/00
• CPC Class Codes: G06F 40/58 Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation

Abstract

A method and system for speech to speech translation using a plurality of computational cores of a natural liquid architecture are provided. The method includes receiving an audio input to be translated into a target language; identifying, based on the received audio input, a cluster of signatures; identifying, based on the cluster of signatures, a first concept, wherein the first concept is common to all signatures within the signature cluster; and determining a match between the first concept and at least a second concept, wherein the second concept is associated with the target language.

Claims

1. A method for speech to speech translation using a plurality of computational cores of a natural liquid architecture, comprising:

receiving an audio input to be translated into a target language;

identifying, based on the received audio input, a cluster of signatures;

identifying, based on the cluster of signatures, a first concept, wherein the first concept is common to all signatures within the signature cluster;

determining a match between the first concept and at least a second concept, wherein the second concept is associated with the target language;

determining a first language based on the first concept, wherein the match between the first concept and at least a second concept is based on the first language, and wherein determining the match between the first concept and the at least a second concept further comprises;

identifying a first plurality of concepts associated with the first language, wherein the first plurality of concepts is statistically proximate to the first concept and previously matched to a second plurality of concepts, wherein the second plurality of concepts is associated with the target language; and

identifying, based on a threshold, a second concept that is statistically proximate to the second plurality of concepts.

2. The method of claim 1, further comprising:

generating, based on the received audio input, at least one signature by the plurality of computational cores of the natural liquid architecture, wherein the cluster of signatures is identified further based on the generated at least one signature.

3. The method of claim 2, further comprising:

matching the generated at least one signature to at least one other signature, wherein the cluster of signatures is identified further based on the matching between the generated at least one signature and the at least one other signature.

4. The method of claim 2, each computational core of the plurality of computational cores having properties that are at least partly statistically independent of other of the computational cores, wherein the properties are set independently of each other core.

5. The method of claim 1, wherein a plurality of concepts is associated with a language when metadata of the plurality of concepts indicates the language.

6. The method of claim 1, wherein the identification of matching concepts is performed by the plurality of computational cores of the natural liquid architecture.

7. The method of claim 1, wherein each concept is at least one of:

at least one tone, at least one musical note, at least one word, and at least one syllable.

8. A non-transitory computer readable medium having stored thereon instructions for causing one or more processing units to execute a method, the method comprising:

receiving an audio input to be translated into a target language;

identifying, based on the received audio input, a cluster of signatures;

identifying, based on the cluster of signatures, a first concept, wherein the first concept is common to all signatures within the signature cluster;

determining a match between the first concept and at least a second concept, wherein the second concept is associated with the target language;

determining a first language based on the first concept, wherein the match between the first concept and at least a second concept is based on the first language, and wherein determining the match between the first concept and the at least a second concept further comprises;

identifying a first plurality of concepts associated with the first language, wherein the first plurality of concepts is statistically proximate to the first concept and previously matched to a second plurality of concepts, wherein the second plurality of concepts is associated with the target language; and

identifying, based on a threshold, a second concept that is statistically proximate to the second plurality of concepts.

9. A system for speech to speech translation, comprising:

a processing circuitry;

an interface coupled to the processing circuitry, wherein the interface is configured to receive an audio signal to be translated into a target language; and

a memory communicatively connected to the processing circuitry, wherein the memory contains instructions that, when executed by the processing circuitry, configure the processing element to;

identify, based on the received audio input, a cluster of signatures;

identify, based on the cluster of signatures, a first concept, wherein the first concept is common to all signatures within the signature cluster; and

determine a match between the first concept and at least a second concept, wherein the second concept is associated with the target language;

determine a first language based on the first concept, wherein the match between the first concept and at least a second concept is based on the first language, and wherein determine the match between the first concept and the at least a second concept further comprises;

identify a first plurality of concepts associated with the first language, wherein the first plurality of concepts is statistically proximate to the first concept and previously matched to a second plurality of concepts, wherein the second plurality of concepts is associated with the target language; and

identify, based on a threshold, a second concept that is statistically proximate to the second plurality of concepts.

10. The system of claim 9, the system including a plurality of computational cores of a natural liquid architecture, wherein the system is further configured to:

generate, based on the received audio input, at least one signature by the plurality of computational cores of the natural liquid architecture, wherein the cluster of signatures is identified further based on the generated at least one signature.

11. The system of claim 10, wherein the system is further configured to:

match the generated at least one signature to at least one other signature, wherein the cluster of signatures is identified further based on the matching between the generated at least one signature and the at least one other signature.

12. The system of claim 10, each computational core of the plurality of computational cores having properties that are at least partly statistically independent of other of the computational cores, wherein the properties are set independently of each other core.

13. The system of claim 9, wherein a plurality of concepts is associated with a language when metadata of the plurality of concepts indicates the language.

14. The system of claim 9, wherein the identification of matching concepts is performed by the plurality of computational cores of the natural liquid architecture.

15. The system of claim 9, wherein each concept is at least one of:

at least one tone, at least one musical note, at least one word, and at least one syllable.