Patent Application
20150073799
A voice verifying system is always applied to a portable electronic apparatus such as a GPS or a mobile phone, such that a user can use a voice command (i.e. a speech signal) to activate the portable electronic device or to control the portable electronic device to perform some function. However, in order to receive the voice, a microphone of the portable electronic device needs to be always turned on to receive the voice. Also, a verifying module of the portable electronic device is always turned on in order to verify the received voice. Therefore, the power consumption for the voice verifying system of related art is high.
One objective of the present application is to provide a voice verifying system that can adjust the standard for determining if the voice signal is valid and turns on at least one device according to the voice signal.
Another objective of the present application is to provide a voice verifying method that can adjust the standard for determining if the voice signal is valid and turns on at least one device according to the voice signal.
One embodiment of the present invention discloses a voice verifying system, which comprises: a microphone, which is always turned on to output at least one voice signal; a speech determining device, for determining if the voice signal is valid or not according to a reference value, wherein the speech determining device passes the voice signal if the voice signal is valid; and a verifying module, for verifying a speech signal generated from the voice signal and for outputting a device activating signal to activate a target device if the speech signal matches a predetermined rule; and a reference value generating device, for generating the reference value according to speech signal information from the verifying module.
Another embodiment of the present invention discloses a voice verifying method, comprising: always turning on a microphone to output at least one voice signal; determining if the voice signal is valid or not according to a reference value, and passing the voice signal to be a speech signal if the voice signal is valid; and verifying the speech signal generated from the voice signal by a predetermined rule, and outputting a device activating signal to activate a target device if the speech signal matches the predetermined rule; generating the reference value according to speech signal information generated from a verifying result.
In view of above-mentioned embodiments, the standard for determining if the voice signal is valid can be adjusted according to at least one factor, and at least one device in the system is turned off (or in the sleep mode) until receives the voice signal. By this way, the power efficiency can be optimized.
These and other objectives of the present invention will no doubt become obvious to those of ordinary skill in the art after reading the following detailed description of the preferred embodiment that is illustrated in the various figures and drawings.
The microphone 101 is always turned on and transmits a voice signal VS to the speech determining device 103. The speech determining device 103 determines if the voice signal VS is valid or not according to a reference value RV, and outputs the voice signal VS if the voice signal VS is valid. For more detail, the microphone 101 may transmits undesired voice signals besides the speech signal SS to the speech determining device 103, and the speech determining device 103 blocks/filters the undesired voice signals and passes the speech signal SS. In one embodiment, at least one of the verifying module 105, the reference value generating device 107 and the environment sensor 109 is initially turned off (or in the sleep mode) and is turned on if the speech determining device 103 outputs the speech signal SS to the verifying module 105. The at least one of the verifying module 105, the reference value generating device 107 and the environment sensor 109 keeps non-active if the speech determining device 103 does not output the speech signal SS to the verifying module 105.
The verifying module 105 receives the speech signal SS and determines if the speech signal SS matches a predetermined rule. Also, the verifying module 105 outputs a device activating signal AS if the speech signal SS matches the predetermined rule and does not output the device activating signal AS if the speech signal SS does not match the predetermined rule. In one embodiment, the predetermined rule is: if the speech signal SS matches a specific command. The specific command can be, for example, “play music”, “play video”, or “call xxx” (xxx can be anyone's name), and the verifying module 105 can be provided in a mobile phone. Alternatively, the predetermined rule can be: if the speech signal SS belongs to a specific person. The voice data for the specific person can be pre-recorded in the verifying module 105 such that the verifying module 105 can identify if the speech signal SS belongs to the specific person. However, please note the predetermined rule is not limited to above-mentioned examples.
As above-mentioned, the verifying module 105 outputs the device activating signal AS if the speech signal SS matches the predetermined rule. The device activating signal AS is utilized to activate a target device 111. The target device 111 can be other devices for the apparatus that the voice verifying system 100 is provided in. For example, the voice verifying system 100 is provided in a mobile phone and the target device 111 is a display which is in a sleep mode until receives the device activating signal AS. Alternatively, in another example the voice verifying system 100 is provided in a GPS apparatus and the target device 111 is a navigating module which is not turned on and does not perform any navigating function until receives the device activating signal AS. Additionally, the target device 111 can be a main processor for the apparatus that the voice verifying system 100 is provided in. Therefore, if the main CPU is initially turned off (or in the sleep mode) and is only activated by the device activating signal AS, the power consumption can be greatly reduced.
The environment sensor 109 senses environment surrounding the environment sensor 109 to generate at least one environment parameter EI to the reference value generating device 107. The reference value generating device 107 generates the reference value RV according to the environment parameter EI. The environment parameter EI comprises at least one of following parameters: a noise level parameter, a motion parameter, a temperature parameter and a location parameter. The noise level parameter indicates the noise situation surrounding the voice verifying system 100. The motion parameter indicates if the voice verifying system 100 moves, and how fast does it move. The temperature parameter indicates the temperature for the environment surrounding the voice verifying system 100. The location parameter indicates the location of the voice verifying system 100.
The noise level may affect the identifying accuracy for the speech signal. Also, the motion parameter or the location parameter can indicate if the user is in the car or in a meeting room, or the user is running, walking, sitting. The temperature may affect the voice signal. Therefore, all these parameters need different standards such that the voice signal can be accordingly determined to be valid or not, therefore the reference value RV should be accordingly adjusted. The environment sensor 109 can be different kinds of sensors to sense the above-mentioned parameters. For example, the environment sensor 109 can comprise a noise sensor to sense the noise level. Also, the environment sensor 109 can comprise a motion sensor such as a G sensor to sense the motion parameter. Alternatively, the environment sensor 109 can comprise an image capture device such as a glass can capture image for identifying the location, thereby generates the location parameter. Additionally, the environment sensor 109 can comprise a temperature sensor to sense the temperature. The above-mentioned sensors can be independent sensors or combined together to form a virtual sensor.
Besides the above mentioned parameters, the reference value generating device 107 may further refer to the microphone sensitivity parameter to generate the reference value RV. The microphone sensitivity parameter indicates the sensitivity of the microphone 101, which affects the voice signal VS. For example, if the microphone sensitivity is high, the microphone outputs a voice signal containing voice for a large region, thus the noise component and the speech component are both high. On the contrary, if the microphone sensitivity is low, the microphone outputs a voice signal containing voice for a small region, thus the noise component and the speech component are both low. Therefore, different standards for determining the voice signals are valid or not are needed. Therefore, the reference value RV should be updated according to the microphone sensitivity parameter.
In one embodiment, the verifying module 105 computes the speech signal SS to acquire speech signal information SSI of the speech signal SS and transmits the speech signal information SSI to the reference value generating device 103. The reference value generating device 103 generates the reference value RV according to the environment parameter EI and the speech signal information SSI. The information of the speech signal comprises at least one of following information: a noise type of the speech signal, a signal to noise ratio (SNR) of the speech signal, a time duration that the speech determining device does not determine that the speech signal exists (i.e. a silent time duration). Either the noise type or the SNR of the speech signal affects the determining of the speech signal. Also, the silent time duration indicates the possibility that the reference value RV is improper for current situation. For example, if the silent time duration is too long, it may indicates the reference value RV is too strict such that the speech signal can not be identified. On the contrary, if the silent time duration is too short, it may indicates the reference value RV is too loose such that the voice signal which is not the speech signal is also wrongly determined to be the speech signal.
In another embodiment, the voice verifying system 100 further comprises a timer 113 to control the time duration that the reference value generating device 107 generates a new reference value RV.
Briefly speaking, the voice verifying system 100 updates the standard (i.e. the reference value RV) for determining if the voice signal VS is valid or not (ex. the voice signal VS is the speech signal SS or not) according to at least one of the environment parameter EI, the speech signal information SSI, and the microphone sensitivity parameter. After that, at least one of the verifying module 105, a reference value generating device 107 and an environment sensor 109 which is initially turned off (or in the sleep mode will be turned on if the speech determining device 103 determines the voice signal VS is valid and outputs the speech signal SS to the verifying module 105. Also, the target device 111 is not activated until receives the device activating signal AS. Therefore, the power consumption for the apparatus utilizing the voice verifying system according to the present application can be reduced.
Many devices can be applied as the speech determining device 303. In the embodiment of
As above-mentioned description, the reference value RV can be generated according to various kinds of parameters, thus the reference value RV can be adjusted corresponding to various situations such that the standard for determining the speech signal can be adjusted to be loose or strict. For example, if the environment parameter EI indicates at least one of following situation the voice verifying system does not move, the voice verifying system is hold in a hand of an user (ex. a mobile phone comprising the voice verifying system is hold in a hand of an user) , and the voice verifying system is in a moving car, then the reference value generating device 107 decreases the reference value RV. If the voice verifying system is in any one of the above-mentioned situations, the user may want to use the voice control function, thus the reference value RV is decreased such that the comparator can easily pass the speech signal to activate other devices. On the contrary, in other situations the user rarely uses the voice control function such that the reference value RV is adjusted to be high thereby the comparator can block most voice signal. By this way other devices can keep non-activated to save power. In another example, the reference value RV is increased if the voice verifying system is put in a pack. The user rarely utilizes the voice control function if the voice verifying system is put in a pack (ex. a mobile phone comprising the voice verifying system is put in a pack) , thus the reference value RV is increased such that other devices can keep non-activated to save power. Please note the above-mentioned situations are only for example and do not mean to limit the scope of the present invention.
In one embodiment, the microphone is a digital microphone, thus the voice signal VS output from the microphone 101 and the speech signal SS are all digital. However, the voice signal VS outputted from the microphone 101 and the speech signal SS can also be an analog signal. In another embodiment, the micro phone is an analog microphone and an analog to digital converter can be further comprised in the voice verifying system
Therefore, the operation for the voice verifying system 300 has some difference from which of the voice verifying system 100 when the microphone 101 is a digital microphone. For the voice verifying system 100, the microphone 101 outputs a digital voice signal VS to the speech determining device 103. The speech determining device 103 determines if the digital voice signal VS is a digital speech signal SS and passes the digital speech signal SS to the verifying module 105 according to the reference value RV.
For the voice verifying system 300, the analog microphone 301 outputs an analog voice signal AVS to the speech determining device 303. The speech determining device 303 determines if the analog voice signal AVS is valid or not according to the reference value RV (ex. if the analog voice signal AVS is an analog speech signal ASS) and passes the analog speech signal ASS to the ADC 313 if the analog voice signal AVS is valid. The ADC 313 converts the analog speech signal ASS to a speech signal SS and transmits the speech signal SS to the verifying module 305.
Additionally, for the voice verifying system 100, at least one of the verifying module 105, a reference value generating device 107 and an environment sensor 109 is initially turned off (or in the sleep mode) and is turned on if the speech determining device 103 outputs the digital speech signal SS to the verifying module 105. For the voice verifying system 300, at least one of the verifying module 305, the reference value generating device 307 and the environment sensor 309 and the ADC 313 is initially turned off (or in the sleep mode) and is turned on if the speech determining device 303 outputs the analog speech signal ASS. By this way, the power consumption can be further reduced since the ADC can be initially turned off or in the sleep mode.
The operations for the verifying module 305, the reference value generating device 307, and the environment sensor 309 in
As above-mentioned, in another embodiment that the ADC 313 can be located in front of the speech determining device 303. In such embodiment, the analog voice signal received by the analog microphone is transferred to a digital voice signal first via the ADC, thus the speech determining device 303 receives the digital voice signal rather then the analog voice signal AVS described above. Additionally, in such embodiment the ADC can be a low power ADC to save power consumption. In such embodiment, operations of the speech determining device, the verifying module , the reference value generating device and the environment sensor are the same as which of
In view of above-mentioned embodiments, a voice verifying method applied to a voice verifying system as shown in
Step 501
Always turn on a microphone to output at least one voice signal .
Step 503
Determine if the voice signal is valid or not according to a reference value. If yes, go to step 505. If not, go to step 515 to block the voice signal.
Step 505
Pass the voice signal to a verifying module.
Step 507
Utilize the verifying module to determine if the speech signal matches a predetermined rule, and utilizing the verifying module to output a device activating signal for activating a target device if the speech signal matches the predetermined rule.
Step 509
Utilize an environment sensor to sense environment surrounding the environment sensor to generate at least one environment parameter.
Step 511
Utilize a reference value generating device to generate the reference value according to the environment parameter.
Step 513
Initially turn off at least one of the verifying module, the environment sensor and the reference value generating device and turn on at least one of the verifying module, the environment sensor and the reference value generating device if the voice signal is determined to be valid and is passed in the previous steps. (or control them to be in the sleep mode) . The at least one of the verifying module, the environment sensor and the reference value generating device keeps non active if the verifying module does not receive the speech signal .
Step 515
Block/filter the voice signal.
In view of above-mentioned embodiments, the standard for determining if the speech signal exists can be adjusted according to at least one factor, and at least one device in the system is turned off until receives the speech signal . By this way, the power efficiency can be optimized.
Those skilled in the art will readily observe that numerous modifications and alterations of the device and method may be made while retaining the teachings of the invention. Accordingly, the above disclosure should be construed as limited only by the metes and bounds of the appended claims.
