Voice-controlled information exchange platform, such as for providing information to supplement advertising

Patent metadata

Field	Value
Publication number	US-9639854-B2
Application number	US-201414316663-A
Country	US
Kind code	B2
Filing date	Jun 26, 2014
Priority date	Jun 26, 2014
Publication date	May 2, 2017
Grant date	May 2, 2017

Abstract

A system with an associated method for preloading advertisements by a server to a user's device is disclosed. In response to inquiries made by members of a user group, the system presents advertisements to the members and keeps a record of these presentations. Next, the system identifies those advertisements which have been frequently presented to the members, and preloads the identified advertisements on the device of a user who belongs to the user group. Subsequently, upon receiving a specific inquiry from the user's device, the system determines a response to the specific inquiry. When the determined response contains one of the preloaded advertisements, the system sends an instruction to the user's device to present the preloaded advertisement to the user.

First claim

The invention claimed is:

1. A computer readable medium, excluding transitory propagating signals, storing instructions that, when executed by at least one computing device, cause the computing device to processing audio signals, comprising:
receiving audio samples at a server from each of multiple users,
wherein the audio samples include ambient noise associated with an environment or location for each of the multiple users, and
wherein at least some of the audio samples are obtained from multiple wireless mobile devices associated with at least some of the multiple users;

selecting at least some of the received audio samples to be processed;
dynamically processing the selected audio samples to produce at least a first processed noise sample for characterizing sounds at the environment or location in real-time,
wherein the first processed noise sample is produced using one or more server processors;

after the processing, receiving an utterance signal from one of the wireless mobile devices associated with one of the multiple users,
wherein the utterance signal includes content and noise; and

filtering the noise from the utterance signal to isolate the content using the first processed noise sample.

2. The computer readable medium of claim 1 ,
wherein receiving the audio samples includes receiving a time and a geographic location associated with each audio sample,
wherein the selecting includes selecting only those audio samples within a time threshold, a geographic threshold or both, and
wherein the processing includes combining the selected audio samples to produce a normalized noise sample.

3. The computer readable medium of claim 1 , wherein the wireless mobile devices include a phone, a watch, eyewear, a tablet computer, a laptop computer, or a personal digital assistant.

4. A method of creating noise signatures for multiple types of environments, comprising:
receiving at a server main segments each containing a foreground sound and a background sound,
wherein the main segments represent audio samples recorded by one or more devices located in one type of environment;

dynamically building a noise signature according to the main segments,
wherein the noise signature characterizes the background sound for the one type of environment in real-time, and
wherein the noise signature is built using one or more server processors;

accepting first audio data containing a first foreground sound and a first background sound for representing another audio sample from another device located in the one type of environment,
wherein the other device is separate from the one or more devices corresponding to the main segments; and

isolating the first foreground sound from the first audio data using the noise signature.

5. The method of claim 4 , further comprising:
determining a voice for presenting a response to the first foreground sound based on the dynamically built instance of the noise signature specifically or contemporaneously corresponding to the first background sound when the dynamically built instance of the noise signature is available; and
determining a voice for presenting a response to the first foreground sound based on a previously built instance of the noise signature for the first type of environments when the main segments specifically or contemporaneously corresponding to the first background sound are unavailable,
wherein the first foreground sound corresponds to a user inquiry.

6. The method of claim 5 , further comprising, when the audio data containing exclusively the first background sound is available and the first background sound is louder than a predetermined volume, alerting the user that the response to the first foreground sound will not be presented until the user is in an environment with a second background sound which is not louder than the predetermined volume by a predetermined amount.

7. The method of claim 4 , wherein the first foreground sound corresponds to a song.

8. The method of claim 4 , further comprising recording an extended segment containing a background sound before the foreground sound begins or after the foreground sound ends.

9. The method of claim 4 , further comprising:
accepting second audio data containing a second background sound in the environment of a second type; and
when a difference between the second background sound and the noise signature for the second type of environments exceeds a predetermined amount, alerting a user to a potential anomaly in the environment of the second type.

10. A system for audio processing for multiple types of environments, comprising:
a communication circuit of a server configured to:
receive main segments each containing a foreground sound and a background sound,
wherein the main segments represent audio samples recorded by one or more devices located in one of multiple types of environments, and

receive an extended segment containing the background sound before the foreground sound begins or after the foreground sound ends; and

one or more server processors, coupled to the communication circuit, configured to:
dynamically build a noise signature for one of the multiple types of environments from the main segments and the extended segments,
wherein the noise signature characterizes the background sound for the one type of environment in real-time,

accept first audio data containing a first foreground sound and a first background sound in an environment of a first type,
isolate the first foreground sound from the first audio data using dynamically built instance of the noise signature specifically or contemporaneously corresponding to the first background sound when the main segments specifically or contemporaneously corresponding to the first background sound is available, and
isolate the first foreground sound from the first audio data using a previously built instance of the noise signature for the first type of environments when the main segments specifically or contemporaneously corresponding to the first background sound are unavailable.

11. The system of claim 10 , wherein the one or more server processors are further configured to:
determine a voice for presenting a response to the first foreground sound based on the dynamically built instance of the noise signature specifically or contemporaneously corresponding to the first background sound when the dynamically built instance of the noise signature is available; and
determine a voice for presenting a response to the first foreground sound based on a previously built instance of the noise signature for the first type of environments when the main segments specifically or contemporaneously corresponding to the first background sound are unavailable,
wherein the first foreground sound corresponds to a user inquiry.

12. The system of claim 11 , wherein the one or more server processors are further configured to generate an alert for the user when the audio data containing exclusively the first background sound is available and the first background sound is louder than a predetermined volume, wherein generating the alert is for notifying the user that the response to the first foreground sound will not be presented until the user is in an environment with a second background sound which is not louder than the predetermined volume by a predetermined amount.

13. The system of claim 10 , wherein the first foreground sound corresponds to a song.

14. The system of claim 10 , wherein the one or more server processors are further configured to:
accept second audio data containing a second background sound in the environment of a second type; and

Assignees

• Nuance Communications Inc

Inventors

• Balasubramanian Sundar
• Mcsherry Michael
• Fu Eric Jun
• Hendrick Daniel
• Katyal Deepankar
• Kay David J

Classifications

• G10L13/033

  Voice editing, e.g. manipulating the voice of the synthesiser · CPC title

• G06Q30/0255 Primary

  based on user history · CPC title

• G10L15/20

  Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech (G10L21/02 takes precedence) · CPC title

• G06Q30/0267

  Wireless devices · CPC title

• G06Q30/0257

  User requested · CPC title