Filtering audio-based interference from voice commands using interference information

Patent metadata

Field	Value
Publication number	US-10832678-B2
Application number	US-201816004234-A
Country	US
Kind code	B2
Filing date	Jun 8, 2018
Priority date	Jun 8, 2018
Publication date	Nov 10, 2020
Grant date	Nov 10, 2020

Abstract

A computer-implemented method, according to one embodiment, includes: receiving a complex audio signal which includes an intended audio signal and at least one interfering audio signal. Moreover, the intended audio signal is a voice-based command originating from a user. Information which corresponds to the at least one interfering audio signal is also received. The received information is used to identify portions of the complex audio signal as being the at least one interfering audio signal. Furthermore, the identified portion of the complex audio signal is removed from the complex audio signal, and a remaining portion of the complex audio signal is output.

First claim

What is claimed is:

1. A computer-implemented method, comprising:
receiving a complex audio signal, wherein the complex audio signal includes an intended audio signal and at least one interfering audio signal, wherein the intended audio signal is a voice-based command originating from a user;
receiving information which corresponds to the at least one interfering audio signal, wherein the received information includes network activity experienced by a network router which corresponds to the user;
using the received information to identify portions of the complex audio signal as being the at least one interfering audio signal;
removing the identified portion of the complex audio signal from the complex audio signal; and
outputting a remaining portion of the complex audio signal,
wherein using the received information to identify portions of the complex audio signal as being the at least one interfering audio signal includes:
identifying audio signals included in the network activity;
comparing the audio signals with the complex audio signal; and

identifying matches between the audio signals and the complex audio signal as portions of the at least one interfering audio signal.

2. The computer-implemented method of claim 1 , wherein the received information includes:
a timestamp which identifies a time that the voice-based command was originated;
content which corresponds to the at least one interfering audio signal; and
an offset which identifies a portion of the content which produced the at least one interfering audio signal at the time that the voice-based command was originated.

3. The computer-implemented method of claim 2 , wherein the content includes an actual copy of the at least one interfering audio signal, or a live stream of the at least one interfering audio signal.

4. The computer-implemented method of claim 1 , wherein the received information includes:
a timestamp which identifies a time that the voice-based command was originated.

5. The computer-implemented method of claim 1 , wherein the received information includes:
a timestamp which identifies a time that the voice-based command was originated; and
audio samples collected from one or more other users at about the time that the voice-based command was originated, the one or more other users being located in a same area as the user which originated the voice-based command,
wherein using the received information to identify portions of the complex audio signal as being the at least one interfering audio signal includes:
comparing the audio samples with the complex audio signal; and
identifying matches between the audio samples and the complex audio signal as portions of the at least one interfering audio signal.

6. The computer-implemented method of claim 1 , wherein using the received information to identify portions of the complex audio signal as being the at least one interfering audio signal includes:
converting the complex audio signal into a plurality of fingerprints;
comparing the plurality of fingerprints to a database of fingerprints which represent known audio signals; and
using matches between the plurality of fingerprints and the database of fingerprints to determine whether the respective portions of the complex audio signal correspond to the interfering audio signal.

7. The computer-implemented method of claim 6 , comprising:
using the received information which corresponds to the at least one interfering audio signal to reduce a size of the database compared to the plurality of fingerprints,
wherein the received information includes contextual information which corresponds to the user and/or user integrated devices in a same environment as the user.

8. The computer-implemented method of claim 1 , wherein the received information includes a second complex audio signal, wherein the second complex audio signal includes the intended audio signal and at least a second interfering audio signal, wherein using the received information to identify portions of the complex audio signal as being the at least one interfering audio signal includes:
comparing the complex audio signal with the second complex audio signal;
identifying matches between the complex audio signal and the second complex audio signal as portions of the intended audio signal; and
identifying the remaining portions of the complex audio signal and the second complex audio signal which do not match as portions of the at least one interfering audio signal.

9. A computer-implemented method, comprising:
receiving a complex audio signal, wherein the complex audio signal includes an intended audio signal and at least one interfering audio signal, wherein the intended audio signal is a voice-based command originating from a user;
converting the complex audio signal into text which represents a plurality of words included in the complex audio signal;
receiving information which corresponds to the at least one interfering audio signal, wherein the received information includes audio samples collected from one or more other users at about a time that the voice-based command was originated, the one or more other users being located in a same area as the user which originated the voice-based command, wherein the audio samples collected from the one or more other users includes the at least one interfering audio signal;
using the received information to identify at least some of the text as representing words which correspond to the at least one interfering audio signal;
discarding the identified text; and
outputting a remaining portion of the text,
wherein the received information includes a second complex audio signal,
wherein the second complex audio signal includes the intended audio signal and at least a second interfering audio signal,
wherein using the received information to identify portions of the complex audio signal as being the at least one interfering audio signal includes:
converting the second complex audio signal into supplemental text which represents a plurality of words included in the second complex audio signal,
comparing the text with the supplemental text,
identifying matches between the text and the supplemental text as representing words which correspond to the intended audio signal, and
identifying the remaining portions of the text and the supplemental text which do not match as representing words which correspond to the at least one interfering audio signal.

10. The computer-implemented method of claim 9 , wherein the received information includes:
a timestamp which identifies the time that the voice-based command was originated;
content which corresponds to the at least one interfering audio signal; and
an offset which identifies a portion of the content which produced the at least one interfering audio signal at the time that the voice-based command was originated.

11. The computer-implemented method of claim 10 , wherein the content includes an actual copy of the at least one interfering audio signal, or a live stream of the at least one interfering audio signal.

12. The computer-implemented method of claim 9 , wherein the received information includes:
a timestamp which identifies the time that the voice-based command was originated; and
network activity experienced by a network router which corresponds to the user,
wherein using the received information to identify at least some of the text as representing words which correspond to the at least one interfering audio signal includes:
identifying audio signals included in the network activity;
converting the identified audio signals into supplemental text which represents a plurality of words includ

Assignees

• IBM

Inventors

• Liu Su
• Rozner Eric J
• Hwang Inseok
• Yoo Chungkuk

Classifications

• G10L21/0208 Primary

  Noise filtering · CPC title

• G10L25/51

  for comparison or discrimination · CPC title

• G10L2015/223

  Execution procedure of a spoken command · CPC title

• G10L15/22

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

• G06F16/90335

  Query processing · CPC title