Prevention of unintended distribution of audio information

• US 9,607,630 B2
• Filed: 04/16/2013
• Issued: 03/28/2017
• Est. Priority Date: 04/16/2013
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Basson, Sara H., Kanevsky, Dimitri, Malkin, Peter K., Wegman, Mark N.
• Primary Examiner(s): ADESANYA, OLUJIMI A
• Application Number:
  US13/863,703
• Publication Number:
  US 20140309998A1
• Time in Patent Office: 1,442 Days
• Field of Search: 704270-275
• US Class Current: 1/1
• CPC Class Codes: G10L 25/00 Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 muting semiconductor-based amplifiers when some special characteristics of a signal are sensed by a speech detector, e.g. sensing when no signal is present, H03G3/34
• CPC Class Codes: G10L 25/51 for comparison or discrimination
• CPC Class Codes: H04R 27/00 Public address systems circuits for preventing acoustic reaction H04R3/02; circuits for distributing signals to loudspeakers H04R3/12; monitoring or testing arrangements for public address systems H04R29/007; amplifiers H03F

Abstract

Preventing unintended distribution of audio information may comprise analyzing audio data of a speaker'"'"'s speech received by a microphone; determining automatically by a processor, from the analyzing whether the speaker'"'"'s speech is intended to be distributed to an audience via the microphone; and in response to determining that the speaker'"'"'s speech is not intended to be distributed to the audience via the microphone, performing one or more actions.

Claims

1. A method of preventing unintended distribution of audio information, comprising:

analyzing, by a processor, audio data of a speaker'"'"'s speech received by a microphone before the audio data is distributed via the microphone;

determining automatically, by the processor, from the analyzing whether the speaker'"'"'s speech is intended to be distributed to an audience via the microphone; and

in response to determining that the speaker'"'"'s speech is not intended to be distributed to the audience via the microphone, performing one or more actions,

wherein the analyzing comprises at least detecting a change in a manner of the speaker'"'"'s speech comprising at least a change in harmonics of the speech, and based on the detecting of the change, determining that the speaker'"'"'s speech is not intended to be distributed to the audience via the microphone,

wherein one or more actions comprises at least generating a signal automatically by the processor that indicates the microphone is turned on, the signal comprising echoing back to the speaker in different harmonics while the speaker is speaking, and turning off the microphone automatically,

wherein the analyzing further comprises detecting a change in a voice volume of a speaker making the speech, a change in a topic of the speech, the method further comprising collecting visual cues comprising a change in distance between a speaker making the speech and the microphone, a change in location from where the speaker is making the speech, the audience entering into and exiting from the location,

the detected change in the voice volume, the change in the topic of the speech, and the visual cues increasing a confidence score in determining to turn off the microphone.

2. The method of claim 1, wherein the analyzing comprises detecting a change in a voice volume of a speaker making the speech, a change in a manner of the speech, a change in a topic of the speech, or combinations thereof.

3. The method of claim 1, further comprising collecting visual cues and the visual cues are also used to determine whether the speaker'"'"'s speech is intended to be distributed.

4. The method of claim 3, wherein the visual cues comprises a change in distance between a speaker making the speech and the microphone, a change in location from where the speaker is making the speech, or combinations thereof.

5. The method of claim 1, further comprising collecting motion data associated with a speaker making the speech, and further using the motion data to determine whether the speaker'"'"'s speech is intended to be distributed.

6. The method of claim 1, wherein the one or more actions comprises providing a feedback to the speaker, muting the microphone, turning off the microphone, or combinations thereof.

7. The method of claim 6, wherein the feedback comprises one or more of flashing lamp, tactile signal, audio signal, a transcription of the speech on a display, or combinations thereof.

8. The method of claim 1, further comprising analyzing non-speech information to determine whether the speaker'"'"'s speech is intended to be distributed.

9. The method of claim 1, wherein the detecting of the change in the manner of the speaker'"'"'s speech further comprises detecting an absence or presence of a filler word in the speaker'"'"'s speech.