Wake suppression for audio playing and listening devices
AssigneeSoundHound, Inc.
InventorsHsuan Yang, Qìndí Zhäng, Warren S. Heit
Life of this patent
Estimating the term…
Abstract
A system and method are disclosed for ignoring a wakeword received at a speech-enabled listening device when it is determined the wakeword is reproduced audio from an audio-playing device. Determination can be by detecting audio distortions, by an ignore flag sent locally between an audio-playing device and speech-enabled device, by and ignore flag sent from a server, by comparison of received audio played audio to a wakeword within an audio-playing device or a speech-enabled device, and other means.
Claims23
A speech-enabled device configured to listen for a wakeword, comprising:
- a microphone configured to receive audio;
- and
- a processor configured to execute software code to:
- receive a first audio signal from the microphone including the wakeword,
- switch the speech-enabled device from an idle state to an active state in response to detecting the wakeword in the first audio signal,
- receive a second audio signal from the microphone including the wakeword,
- determine that the second audio signal is reproduced audio, and
- leave the speech-enabled device in the idle state in response to detecting the second wakeword in the second audio signal and determining that the second audio signal is reproduced audio.
A first device having a connection to a second, speech-enabled device, the speech-enabled device including a microphone and the speech-enabled device configured to listen for a wakeword, the first device comprising:
- a network interface;
- and
- a processor configured to execute software code to:
- analyze audio to be broadcast to a microphone of the speech-enabled device,
- spot, in the audio to be broadcast, a wakeword known to be used to activate the speech-enabled device, and
- transmit, upon spotting the wakeword, an ignore signal to the speech-enabled device configured to have the speech-enabled device ignore the wakeword upon receipt of the wakeword at the microphone of the speech-enabled device.
A speech-enabled device having a connection to an audio-playing device, the speech-enabled device configured to listen for a wakeword, the speech-enabled device comprising:
- a processor configured to execute software code to:
- receive audio, including the wakeword, from the audio-playing device,
- analyze the received audio to spot the wakeword, and
- ignore the wakeword spotted in the audio, to leave the speech-enabled device in an idle state, upon determining the audio came from the audio-playing device.
A speech-enabled device having a connection to an audio-playing device, the speech-enabled device configured to listen for a wakeword, the speech-enabled device comprising:
- a microphone configured to receive audio;
- and
- a processor configured to execute software code to:
- receive a stream of audio, including the wakeword, from the audio-playing device via the connection,
- analyze the received audio stream to spot the wakeword,
- receive a broadcast of the audio at the microphone, including the wakeword, from the audio-playing device,
- analyze the received audio broadcast to spot the wakeword,
- ignore the wakeword spotted in the broadcast of the audio received at the microphone to leave the speech-enabled device in an idle state.
A speech-enabled device configured to listen for a wakeword, comprising:
- a microphone configured to receive audio;
- and
- means for ignoring the wakeword received at the microphone, to leave the speech-enabled device in an idle state, upon determining the audio is reproduced audio.
A device for producing a media stream based on original audio content provided in an audio elementary stream the audio content having speech of a wakeword at an identifiable time, comprising:
- a stream formatter configured to generate a second elementary stream with a signal of the time of the wakeword within the original audio content;
- and
- a server configured to provide the audio elementary stream and the second elementary stream to a playing device;
- wherein said stream formatter is configured to detect the wakeword in the audio elementary stream, identify its time position, and encode the time position into the second elementary stream to produce a media stream.
Dependencies are inferred from the claim text.