Power-Efficient Voice Activation

Abstract

A voice activation system is provided. The voice activation system includes a first module configured to receive an audio signal and output an activation signal if an energy characteristic of the audio signal satisfies a threshold stored in a memory, a control module configured to enable or disable a third state using a control signal, and a speech recognition engine coupled to the first module and the control module, the speech recognition engine configured to transition between a first state, a second state, and the third state. The speech recognition engine transitions from the first state to the second state in response to the activation signal and in response to the control signal being disabled. The speech recognition engine transitions from the first state to the third state in response to the activation signal and in response to the control signal being enabled. The speech recognition engine transitions from the third state to the second state in response to detection of a wake-up word by the speech recognition engine.

1 - 20 . (canceled) 
     
     
         21 . A voice activation system, comprising:
 a first module configured to receive an audio signal and output an activation signal if an energy characteristic of the audio signal satisfies a threshold stored in a memory;   a control module configured to enable or disable a third state using a control signal; and   a speech recognition engine coupled to the first module and the control module, the speech recognition engine configured to transition between a first state, a second state, and the third state, wherein the speech recognition engine is configured to transition from the first state to the second state in response to the activation signal and in response to the control signal being disabled, wherein the speech recognition engine is configured to transition from the first state to the third state in response to the activation signal and in response to the control signal being enabled, wherein the speech recognition engine is configured to transition from the third state to the second state in response to detection of a wake-up word by the speech recognition engine, wherein the second state of the speech recognition engine is configured to consume less power than the third state of the speech recognition engine, and wherein the first module and speech recognition engine are implemented in a computing device of the voice activation system.   
     
     
         22 . The voice activation system of  claim 21 , further comprising a second module configured to transition from a first state of the second module to a second state of the second module in response to a first activation signal from the first module and to output a second activation signal to the speech recognition engine, and wherein the speech recognition engine is to receive the second activation signal as the activation signal in the first state of the speech recognition engine. 
     
     
         23 . The voice activation system of  claim 22 , wherein the second module is configured to compute one or both of a time and frequency profile of the audio signal when in the second state of the second module, and output the second activation signal if at least a portion of the computed time or frequency profile of the audio signal substantially matches at least one predetermined time or frequency profile indicative of speech in the audio signal, and wherein the first module is configured to consume less power than the second module. 
     
     
         24 . The voice activation system of  claim 21 , wherein the speech recognition engine is configured to transition from the third state to the first state in response to the wake-up word not being recognized in the audio signal by the speech recognition engine. 
     
     
         25 . The voice activation system of  claim 21 , wherein the first state is a standby state, the second state is a fully-operational state, and the third state is a wake-up word determination state, wherein the speech recognition engine is configured to receive the audio signal and determine a wake-up word in the audio signal while in the wake-up word determination state, and wherein the speech recognition engine is configured to receive the audio signal and determine a speech content of the received audio signal in the fully-operational state. 
     
     
         26 . The voice activation system of  claim 21 , wherein the control module is configured to enable the speech recognition engine to transition to the wake-up word detection state using the control signal based on at least input from a user. 
     
     
         27 . The voice activation system of  claim 21 , wherein the first module is configured to compare an energy level of the audio signal to the threshold. 
     
     
         28 . The voice activation system of  claim 21 , wherein the first module is configured to compare a ratio of high frequency energy to low frequency energy in the audio signal to the threshold. 
     
     
         29 . A method comprising:
 comparing at least one energy characteristic of an audio signal to a threshold using a first module of a voice activation system;   outputting, by the first module, an activation signal when an energy characteristic of the audio signal satisfies a threshold stored in memory;   outputting, by a control module of the voice activation system, a control signal to enable or disable a third state of a speech recognition engine; and   transitioning the speech recognition engine of the voice activation system between a first state, a second state, and a third state, wherein the transitioning comprises:
 transitioning from the first state to the second state in response to the activation signal and in response to the control signal being disabled; 
 transitioning from the first state to the third state in response to the activation signal and in response to the control signal being enabled; and 
 transitioning from the third state to the second state in response to detection of a wake-up word by the speech recognition engine, wherein the second state of the speech recognition engine is configured to consume less power than the third state of the speech recognition engine. 
   
     
     
         30 . The method of  claim 29 , wherein:
 the transitioning from the first state to the second state further comprises transitioning a speech recognition engine of the voice activation system from a first state to a second state if the least a portion of a profile of the audio signal substantially matches the at least one predetermined profile and if a wake-up word determination state is disabled; and   the transitioning from the first state to the third state further comprises transitioning the speech recognition engine from the first state to the third state if the portion of the profile of the audio signal substantially matches the at least one pre-determined profile and if the wake-up word determination state is enabled.   
     
     
         31 . The method of  claim 29 , further comprising:
 transitioning, by a second module of the voice activation system, from a first state of the second module to a second state of the second module in response to a first activation signal from the first module; and   outputting, by the second module, a second activation signal to the speech recognition engine.   
     
     
         32 . The method of  claim 31 , further comprising:
 computing, by the second module in the second state, one or both of a time and frequency profile of the audio signal; and   comparing, by the second module in the second state, at least a portion of one or both of the computed time and frequency profiles of the audio signal to at least one predetermined time or frequency profile indicative of speech in the audio signal, and wherein the first module consumes less power than the second module, and wherein the second module consumes less power than the third state of the speech recognition engine.   
     
     
         33 . The method of  claim 32 , wherein comparing at least one energy characteristic of the audio signal comprises comparing an energy level of the audio signal to the threshold. 
     
     
         34 . The method of  claim 32 , wherein comparing at least one energy characteristic of the audio signal comprises comparing a ratio of high frequency energy to low frequency energy in the audio signal to the threshold. 
     
     
         35 . The method of  claim 32 , wherein comparing at least the portion of the profile of the audio signal comprises:
 extracting a feature of the audio signal; and   comparing the feature to the at least one predetermined profile.   
     
     
         36 . The method of  claim 29 , wherein the second state of the speech recognition engine is a fully operational state. 
     
     
         37 . The method of  claim 29 , wherein the third state of the speech recognition engine is a wake-up word detection state, the method further comprising:
 determining whether at least one wake-up word is present in the audio signal using the speech recognition engine while the speech recognition engine is in the wake-up word detection state; and   transitioning the speech recognition engine from the wake-up word detection state to the second state if the at least one wake-up word is present in the audio signal, wherein the second state is a fully operational state of the speech recognition engine.   
     
     
         38 . The method of  claim 37 , further comprising enabling the speech recognition engine to transition to the wake-up word detection state using the control signal. 
     
     
         39 . A voice activation system, comprising:
 a microphone configured to output an analog electrical signal corresponding to received sound waves;   an analog-to-digital converter configured to covert the analog electrical signal to a digital signal;   a first module configured to output a first activation signal if at least one energy characteristic of the digital signal satisfies at least one threshold;   a control module configured to enable or disable a third state using a control signal; and   a speech recognition engine coupled to the first module and the control module, the speech recognition engine configured to transition between a first state, a second state, and the third state, wherein the speech recognition engine is configured to transition from the first state to the second state in response to the activation signal and in response to the control signal being disabled, wherein the speech recognition engine is configured to transition from the first state to the third state in response to the activation signal and in response to the control signal being enabled, wherein the speech recognition engine is configured to transition from the third state to the second state in response to detection of a wake-up word by the speech recognition engine, wherein the second state of the speech recognition engine is configured to consume less power than the third state of the speech recognition engine, and wherein the first module and speech recognition engine are implemented in a computing device of the voice activation system.   
     
     
         40 . The voice activation system of  claim 39 , further comprising a second module configured to transition from a first state of the second module to a second state of the second module in response to a first activation signal from the first module and to output a second activation signal to the speech recognition engine, and wherein the speech recognition engine is to receive the second activation signal as the activation signal in the first state of the speech recognition engine.