Low power activation of a voice activated device

• US 9,043,211 B2
• Filed: 05/08/2014
• Issued: 05/26/2015
• Est. Priority Date: 05/09/2013
• Status: Active Grant

• Current Assignee: DSP Group Ltd. (Synaptics Incorporated)
• Original Assignee: DSP Group Ltd. (Synaptics Incorporated)
• Inventors: Haiut, Moshe, Heiman, Arie, Yehuday, Uri
• Primary Examiner(s): Harper, Vincent P
• Application Number:
  US14/272,544
• Publication Number:
  US 20140337036A1
• Time in Patent Office: 383 Days
• Field of Search: 704/275, 704/251, 704/270
• US Class Current: 704/275
• CPC Class Codes: G06F 1/3265 Power saving in display device
• CPC Class Codes: G06F 3/167 Audio in a user interface, e.g. using voice commands for navigating, audio feedback
• CPC Class Codes: G10L 15/20 Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech G10L21/02 takes precedence
• CPC Class Codes: G10L 15/28 Constructional details of speech recognition systems
• CPC Class Codes: G10L 25/78 Detection of presence or absence of voice signals switching of direction of transmission by voice frequency in two-way loud-speaking telephone systems H04M9/10

Abstract

In a mobile device, a bone conduction or vibration sensor is used to detect the user'"'"'s speech and the resulting output is used as the source for a low power Voice Trigger (VT) circuit that can activate the Automatic Speech Recognition (ASR) of the host device. This invention is applicable to mobile devices such as wearable computers with head mounted displays, mobile phones and wireless headsets and headphones which use speech recognition for the entering of input commands and control. The speech sensor can be a bone conduction microphone used to detect sound vibrations in the skull, or a vibration sensor, used to detect sound pressure vibrations from the user'"'"'s speech. This VT circuit can be independent of any audio components of the host device and can therefore be designed to consume ultra-low power. Hence, this VT circuit can be active when the host device is in a sleeping state and can be used to wake the host device on detection of speech from the user. This VT circuit will be resistant to outside noise and react solely to the user'"'"'s voice.

Claims

1. A voice activated device, comprising:

a voice trigger circuit,

an automatic speech recognition circuit;

a controller;

a set of detectors;

wherein the voice trigger circuit is arranged to receive first detection signals from a first sensor of the set of detectors and to trigger an activation of the automatic speech recognition circuit in response to a detection of a voice trigger, while the automatic speech recognition device is in a first power mode;

wherein the automatic speech recognition circuit is arranged to detect, while operating in a second power mode, an audio activation command;

wherein a power consumption of the automatic speech recognition circuit when operating in the second power mode exceeds a power consumption of the automatic speech recognition when operating in the first power mode;

wherein the first sensor is a bone conduction sensor or a vibration sensor; and

wherein the controller is arranged to determine which sub-set of detectors to activate when the automatic speech recognition device is in the first power mode, in response to previous voice trigger detections and automatic speech recognition detections;

wherein the sub-set of detectors belongs to the set of detectors.

2. The voice activated device according to claim 1 the automatic speech recognition is arranged to be idle or deactivated when in the first power mode.

3. The voice activated device according to claim 1 comprising a display, wherein the first sensor is a vibration sensor that is connected to the display.

4. The voice activated device according to claim 1 comprising a buffer that is arranged to receive audio samples from an audio sensor and to store the audio samples, wherein the automatic speech recognition circuit, after being activated, is arranged to retrieve the audio samples;

wherein at least some audio samples are received by the buffer when the automatic speech recognition device is in the first power mode.

5. The voice activated device according to claim 4 wherein the voice activated device comprises the audio sensor.

6. The voice activated device according to claim 1, wherein the automatic speech recognition circuit is arranged to detect the audio activation command in response to audio samples provided by an audio sensor and in response to the first detection signals.

7. The voice activated device according to claim 1, wherein the voice trigger circuit is further arranged to receive audio samples from an audio sensor and to detect the voice trigger in response to the first detection signals and the audio samples.

8. The voice activated device according to claim 1, wherein the voice trigger circuit is arranged to activate an audio sensor if the first detection signals are indicative of a potential voice trigger;

and to verify the voice trigger if audio samples generated by the audio sensor are indicative of the voice trigger.

9. The voice activated device according to claim 1 wherein the set of detectors comprises the vibration sensor and the bone conduction sensor.

10. The voice activated device according to claim 1 wherein the controller is arranged to determine which sub-set of detectors to activate also when the automatic speech recognition device is in the first power mode, in response to an expected power consumption associated with an activation of the sub-set of detectors.

11. The voice activated device according to claim 1, wherein the voice trigger circuit is further arranged to receive audio samples from an audio sensor and to detect the voice trigger in response to the first detection signals and the audio samples.

12. The voice activated device according to claim 1, wherein the voice trigger circuit is arranged to activate an audio sensor if the first detection signals are indicative of a potential issuance of a voice trigger;

and to verify the voice trigger if audio samples generated by the audio sensor are indicative of the voice trigger.

13. A method for activating a voice activated device, the method comprises:

determining which sub-set of detectors to activate when an automatic speech recognition process is deactivated in response to previous voice trigger detections and automatic speech recognition detections;

wherein the sub-set of detectors belongs to a set of detectors;

generating, by a first sensor of the set of detectors, first detection signals;

wherein the first sensor is a bone conduction sensor or a vibration sensor;

detecting a voice trigger in response to the first detection signals;

triggering the automatic speech recognition process in response to the detecting of the voice trigger; and

detecting, by the automatic speech recognition process, an audio activation command; and

deactivating the speech recognition process;

wherein power consumption associated with the automatic speech recognition process exceeds a power consumption associated with the detecting of the voice trigger.

14. The method according to claim 13 comprising buffering audio samples from an audio sensor;

wherein at least some audio samples are buffer prior to the triggering of the automatic speech recognition process; and

processing the audio samples by the automatic speech recognition process.

15. The method according to claim 13 wherein the automatic speech recognition process is responsive to audio samples provided by an audio sensor and to the first detection signals.