Systems and methods for automatically unmuting conference participants
AssigneeRingCentral, Inc.
InventorAleksandr Kodintsev
Life of this patent
Estimating the term…
Abstract
Disclosed is a conference system and associated methods for automatically controlling the audio in a conference involving multiple participants. The system receives and analyzes the audio streams associated with each of the participants. The system detects an identifier that is mentioned in the audio of a first audio stream, determines a context with which the identifier is mentioned in the audio of the first audio stream, and unmutes a second audio stream in response to the identifier being linked to the second audio stream and further in response to the context from the audio of the first audio stream specifying a request that a user associated with the second audio stream speak.
Claims20
A computer-implemented method for automatically controlling audio in a conference, the computer-implemented method comprising:
- receiving, at a conference system, a plurality of audio streams for the conference;
- detecting an audio sample in one of the plurality of audio streams that indicates a formal start of the conference;
- in response to detecting the audio sample, muting the plurality of audio streams except for a first audio stream of the plurality of audio streams that is associated with a host of the conference;
- detecting, by execution of the conference system, an identifier that is mentioned in an audio of the first audio stream and that identifies a user that is associated with a second audio stream of the plurality of audio streams;
- retrieving context that is spoken before and after the identifier is mentioned in the audio of the first audio stream;
- classifying, by execution of the conference system, the context associated with the identifier as a statement or question that is directed to the user that is identified by the identifier and that invokes a response from the user without including a command or signal to unmute the user or the second audio stream;
- and
- unmuting, by execution of the conference system, the second audio stream in response to classifying the context as the statement or question that invokes the response from the user without including the command or signal to unmute the user or the second audio stream.
A conference system for automatically controlling audio in a conference, the conference system comprising:
- one or more hardware processors configured to:
- receive a plurality of audio streams for the conference;
- detect an audio sample in one of the plurality of audio streams that indicates a formal start of the conference;
- in response to detecting the audio sample, mute the plurality of audio streams except for a first audio stream of the plurality of audio streams that is associated with a host of the conference;
- detect an identifier that is mentioned in an audio of the first audio stream and that identifies a user that is associated with a second audio stream of the plurality of audio streams;
- retrieve context that is spoken before and after the identifier is mentioned in the audio of the first audio stream;
- classify the context associated with the identifier as a statement or question that is directed to the user that is identified by the identifier and that invokes a response from the user without including a command or signal to unmute the user or the second audio stream;
- and
- unmute the second audio stream in response to classifying the context as the statement or question that invokes the response from the user without including the command or signal to unmute the user or the second audio stream.
A non-transitory computer-readable medium storing program instructions that, when executed by one or more hardware processors of a conference system, cause the conference system to perform operations comprising:
- receive a plurality of audio streams for a conference;
- detect an audio sample in one of the plurality of audio streams that indicates a formal start of the conference;
- in response to detecting the audio sample, mute the plurality of audio streams except for a first audio stream of the plurality of audio streams that is associated with a host of the conference;
- detect an identifier that is mentioned in an audio of the first audio stream and that identifies a user that is associated with a second audio stream of the plurality of audio streams;
- retrieve context that is spoken before and after the identifier is mentioned in the audio of the first audio stream;
- classify the context associated with the identifier as a statement or question that is directed to the user that is identified by the identifier and that invokes a response from the user without including a command or signal to unmute the user or the second audio stream;
- and
- unmute the second audio stream in response to classifying the context as the statement or question that invokes the response from the user without including the command or signal to unmute the user or the second audio stream.
Dependencies are inferred from the claim text.