Dialogue system with audio watermark

Patent metadata

Field	Value
Publication number	US-9818414-B2
Application number	US-201514731315-A
Country	US
Kind code	B2
Filing date	Jun 4, 2015
Priority date	Jun 4, 2015
Publication date	Nov 14, 2017
Grant date	Nov 14, 2017

Abstract

Described is an apparatus which comprises: first logic to generate a first audio data and to embed the first audio data with a watermark to generate an embedded data; a speaker to output the embedded data as a first audible audio; a microphone to receive a second audible audio; and second logic to check the second audible audio for the watermark, and if the second audible audio has the watermark embedded in the first audio data, generate a first message, else generate a second message.

First claim

I claim:

1. A machine-readable media is provided having machine executable instructions, that when executed cause one or more processors to perform an operation immune from self-hearing, the operation comprising:
generate a first audio data;
embed the first audio data with a watermark to generate an embedded data;
output the embedded data as a first audible audio via a speaker; and
receive a second audible audio and check the second audible audio for the watermark, and if the second audible audio has the watermark embedded in the first audio data, generate a first message to avoid self-hearing, else generate a second message in response to the second audible audio.

2. The machine-readable media of claim 1 having further machine executable instructions, that when executed cause the one or more processors to perform a further operation comprising:
generate the watermark by applying a phase spectrum of a block of the first audio data.

3. The machine-readable media of claim 2 , wherein the block of the first audio data is a first block.

4. The machine-readable media of claim 1 , wherein the operation to check the second audible audio for the watermark comprises:
extract a watermark from a predetermined block of data of the second audible audio; and
compare the extracted watermark against the watermark embedded in the first audio data.

5. The machine-readable media of claim 1 , wherein the first message indicates that the second audible audio is the same as the first audible audio.

6. The machine-readable media of claim 1 , having further machine executable instructions, that when executed cause the one or more processors to perform a further operation comprising:
perform automatic speech recognition (ASR) on the second audible audio when the second audible audio does not have the watermark.

7. The machine-readable media of claim 1 , wherein the watermark is one of:
a random number of variable length,
a random number of fixed length, or
a prime number.

8. The machine-readable media of claim 1 , wherein the one or more processors are part of a wearable device.

9. The machine-readable media of claim 1 , having further machine executable instructions, that when executed cause the one or more processors to perform a further operation comprising:
maintain a table of watermarks to identify watermarks from audios generated by multiple external sources.

10. The machine-readable media of claim 9 , having further machine executable instructions, that when executed cause the one or more processors to perform a further operation comprising:
generate one or more messages when a received audio includes one of the watermarks in the table.

11. An apparatus comprising:
first logic to generate a first audio data and to embed the first audio data with a watermark to generate an embedded data;
a speaker to output the embedded data as a first audible audio;
a microphone to receive a second audible audio; and
second logic to check the second audible audio for the watermark, and if the second audible audio has the watermark embedded in the first audio data, generate a first message to avoid self-hearing, else generate a second message in response to the second audible audio.

12. The apparatus of claim 11 comprises third logic to generate the watermark by applying a phase spectrum of a block of the first audio data.

13. The apparatus of claim 11 , wherein the second logic to:
extract a watermark from a predetermined block of data of the second audible audio; and
compare the extracted watermark against the watermark embedded in the first audio data.

14. The apparatus of claim 11 , wherein the first message indicates that the second audible audio is the same as the first audible audio.

15. The apparatus of claim 11 comprises a digital signal processing logic to perform automatic speech recognition (ASR) on the second audible audio when the second audible audio does not have the watermark.

16. The apparatus of claim 11 , wherein the watermark is one of:
a random number of variable length,
a random number of fixed length, or
a prime number.

17. A system comprising:
a memory;
a speaker;
a microphone;
a processor coupled to the memory, the processor including:
first logic to generate a first audio data and to embed the first audio data with a watermark to generate an embedded data, wherein the speaker to output the embedded data as a first audible audio, and wherein the microphone to receive a second audible audio; and
second logic to check the second audible audio for the watermark, and if the second audible audio has the watermark embedded in the first audio data, generate a first message to avoid self-hearing, else generate a second message in response to the second audible audio; and

a wireless interface to allow the processor to communicate with another device.

18. The system of claim 17 , wherein the processor comprises: third logic to generate the watermark by applying a phase spectrum of a block of the first audio data.

19. The system of claim 17 , wherein the processor comprises a digital signal processing logic to perform automatic speech recognition (ASR) on the second audible audio when the second audible audio does not have the watermark.

20. The system of claim 17 , wherein the watermark is one of: a random number of variable length, a random number of fixed length, or a prime number.

Assignees

• Intel Corp

Inventors

• Phielipp Mariano J

Classifications

• G10L15/22 Primary

  Procedures used during a speech recognition process, e.g. man-machine dialogue · CPC title

• G10L2015/225

  Feedback of the input speech · CPC title

• G10L19/018 Primary

  Audio watermarking, i.e. embedding inaudible data in the audio signal · CPC title