Patent Application
20130336500
This application claims priority/priorities from Japanese Patent Application No. 2012-138184 filed on Jun. 19, 2012, the entire contents of which are incorporated herein by reference.
Embodiments described herein relate generally to a signal processing apparatus and a signal processing method.
Conventionally, a disturbance component such as a noise component or an echo component which is contained in an audio signal is reduced by correcting the audio signal with a noise canceller, an echo canceller, or the like using a DSP (digital signal processor) or the like.
In particular, among such electronic apparatus as PDAs (personal digital assistants) and cell phones are ones in which noise introduced into an apparatus which is gripped by a user or attached to something is detected and a countermeasure is taken in a direction in which the apparatus is affected.
For example, there is known a method (linear acoustic echo canceller) for eliminating acoustic echo due to air propagation of sound from a speaker. Further, there is known a method (nonlinear acoustic echo canceller) for eliminating acoustic echo (nonlinear component) due to speaker vibration, for example. Still further, there is know a method (double microphone acoustic echo canceller) for eliminating acoustic echo due to air propagation of sound from a speaker using an adaptive filter which uses, as a reference signal, sound that is emitted from the speaker and goes around and reaches microphones.
However, none of the above methods directly take into consideration sound that is emitted from a speaker and goes around and reaches a microphone through body vibration (solid propagation sound) or an echo path variation due to apparatus body motion that is caused by a user action.
That is, although a technique is desired which can suppress echo and noise that are introduced into a microphone through solid propagation (i.e., propagation through an apparatus body) of vibration that originates from a speaker, no means capable of satisfying that desire seems to be known.
A general architecture that implements the various features of the present invention will now be described with reference to the drawings. The drawings and the associated descriptions are provided to illustrate embodiments and not to limit the scope of the present invention.
One embodiment provides a signal processing apparatus, including: a speaker; a vibration sensor; and a controller. The speaker is configured to output a sound. The vibration sensor is configured to detect a vibration that is caused by a solid propagation of the sound from the speaker, and to output a reference signal based on the detected variation. The controller is configured to perform a noise suppress control which suppresses a noise due to the vibration using the reference signal.
Embodiments will be hereinafter described.
An electronic apparatus 100 and its control method according to a first embodiment will be described in detail with reference to
The electronic apparatus 100 has a thin, box-shaped body B and the screen of a display unit 11 is generally flush with the front surface of the body B. The display unit 11 is equipped with a touch panel 111 (see
The display unit 11 is composed of a touch panel 11 and a display 112 such as an LCD (liquid crystal display) or an organic EL (electroluminescence) display. The touch panel 11 can detect a position (touch position) on the display screen where it has been touched by, for example, a finger of the user who is gripping the body B. This function of the touch panel 111 allows the display 112 to serve as what is called a touch screen.
The CPU 12 is a central processor for controlling operations of the electronic apparatus 100, and controls individual components of the electronic apparatus 100 via the system controller 13. The CPU 12 realizes individual functional sections (described below with reference to
The system controller 13 incorporates a memory controller for access-controlling the nonvolatile memory 17 and the RAM 18. The system controller 13 also has a function of performing a communication with the graphics controller 14.
The graphics controller 14 is a display controller for controlling the display 112 which is used as a display monitor of the electronic apparatus 100. The touch panel controller 15 controls the touch panel 111 and thereby acquires, from the touch panel 111, coordinate data that indicates a user touch position on the display screen of the display 112.
For example, the acceleration sensor 16 is a 6-axis acceleration sensor capable of detection of acceleration in and around the three directions shown in
Each vibration sensor 23 converts, inside itself, a signal generated by a vibration sensing element into a digital vibration signal xf[n] (n=1, 2, . . . ) and outputs the latter.
The audio processing section 20 performs audio processing such as digital conversion, noise elimination, and echo cancellation on audio signals supplied from the microphones 21, and outputs a resulting signal to the CPU 12. Furthermore, the audio processing section 20 performs audio processing such as voice synthesis under the control of the CPU 1, and supplies a generated audio signal to the speakers 22 to make a voice notification through the speakers 22.
The audio processing section 20 is accompanied by a volume unit (user volume) 31 and equipped with a D/A converter 32.
The volume unit 31 adjusts the sound volume of an audio signal that is supplied from a communication section 24A via a decoding section 12A according to a manipulation amount of a volume adjustment switch.
The D/A converter 32 converts a digital audio signal xa[n] (n=1, 2, . . . ) as volume-adjusted by the volume unit 31 into an analog signal and outputs the latter to the speakers 22. The speakers 22, which are stereo speakers (alternatively, a monaural speaker is used), output a sound (reproduction sound) to the space in which the electronic apparatus 100 exists. The speakers 22 converts the analog signal supplied from the D/A converter 32 into physical vibration and thereby outputs a sound.
On the other hand, the audio processing section 20 is equipped with an A/D converter 33 which is connected to the microphones 21. The microphones 21, which are stereo microphones (alternatively, a monaural microphone is used), pick up a sound that is traveling through the space where the electronic apparatus 100 exists. The microphones 21 convert the picked-up sound into an analog picked-up sound signal z(t) (t: time) and outputs the latter to the A/D converter 33.
The A/D converter 33 converts the analog picked-up sound signal z(t) into a digital signal z[n] (n=1, 2, . . . ) and outputs the latter to an echo/noise suppressing section 20A which is a controller for suppressing echo and noise. A coding section 12B encodes a digital audio signal as noise-suppressed by the echo/noise suppressing section 20A and outputs a resulting signal to the communication section 24A. The decoding section 12A and the coding section 12B are functions of the CPU 12.
A configuration for acoustic echo elimination which, instead of performing a voice call, makes it possible to perform voice recognition while outputting a sound of a content such as a TV program or music is obtained by replacing the decoding section 12A with a memory (not shown) which is stored with contents of TV programs, music, etc. and replacing the coding section 12B with a voice recognizing section (not shown).
A pulse wave sensor 34 receives a human pulse wave and outputs a corresponding digital signal v[n] (n=1, 2, . . . ) to the vital signal clearing processing section 20B. The vital signal clearing processing section 20B performs vital signal clearing processing using an output of the acceleration sensor 16 (to eliminate noise that results from vibration caused by user motion also from the vital signal) and outputs of the vibration sensors 23 (to eliminate noise that results from vibration produced by the speakers 22 also from the vital signal), and outputs a resulting signal to the communication section 24B. For example, vital signal clearing processing section 20B suppresses noise in a vital signal v[n] by processing the vital signal v[n] with adaptive filter using outputs of the acceleration sensor 16 and the vibration sensors 23 as reference signals. Although in this example a pulse wave is employed as an example vital signal, any of other vital signals such as a pulse, a brain wave, an electrocardiogram, an electromyogram, a body temperature, a heartbeat, a skin surface temperature, a skin potential, a blood volume, a breathing rate, a blood saturation oxygen level (SpO2), and an O2Hb concentration may be used as a vital signal.
The delay buffer 211 adjusts the signal time difference so that the reading of a digital signal xa[n] is timed with introduction, through going-around, of a reproduction sound of the digital signal xa[n] into a digital signal z[n] of a picked-up sound signal. The doubletalk detecting section 212 detects doubletalk using xa[n] and z[n] (or an echo-reduced version of z[n]). The filter coefficients updating section 213 updates filter coefficients according to a detection result of the doubletalk detecting section 212. The filter coefficients updating section 213 does not update the filter coefficients if doubletalk is detected by the doubletalk detecting section 212. The filter coefficients memory 214 holds updated filter coefficients. The pseudo-echo generating section 215 generates pseudo-echo using the updated filter coefficients. The echo reducing section 216 reduces echo on the basis of the generated pseudo-echo. The echo path variation detecting section 217 controls the degree of update of the filter coefficients on the basis of an output of the acceleration sensor 16. If detecting an echo path variation, the echo path variation detecting section 217 increases the degree of update so that the filter coefficients are changed quickly to a large extent.
The delay buffer 221 adjusts the signal time difference so that the reading of a digital signal xf[n] of vibration is timed with introduction, through going-around, of solid vibration caused by outputs of the speakers 22 into a digital signal z[n] of a picked-up sound signal. The doubletalk detecting section 222 detects doubletalk using the digital signal xf[n] of the vibration and the digital signal z[n] (or an echo-reduced version of z[n]). The filter coefficients updating section 223 updates filter coefficients according to a detection result of the doubletalk detecting section 222. The filter coefficients updating section 223 does not update the filter coefficients if doubletalk is detected by the doubletalk detecting section 222. The filter coefficients memory 224 holds updated filter coefficients. The pseudo-echo generating section 225 generates pseudo-echo using the updated filter coefficients. The echo reducing section 226 reduces echo on the basis of the generated pseudo-echo.
Step S81: Delays a reproduction signal xa[n].
Step S82: Detects an echo path variation on the basis of an output of the acceleration sensor 16.
Step S83: Updates filter coefficients ha[n] according to an echo path variation, and generates pseudo-echo on the basis of a delayed version of the signal xa[n].
Step S84: Reduces echo in a picked-up sound signal z[n] using the pseudo echo, and outputs a resulting signal.
Step S85: Delays a signal xf[n] of the vibration sensors 23.
Step S86: Updates filter coefficients hf[n] on the basis of a delayed version of the signal xf[n], and generates pseudo-echo.
Step S87: Reduces echo in the picked-up sound signal z[n] using the pseudo echo, and outputs a resulting signal.
In this process, the filter coefficients ha[n] of the first echo suppressing section 20A1 are updated on the basis of an echo-reduced signal which is an output of the first echo suppressing section 20A1 and the filter coefficients hf[n] of the second echo suppressing section 20A2 are updated on the basis of an echo-reduced signal which is an output of the second echo suppressing section 20A2. That is, the first echo suppression and the second echo suppressed are performed sequentially.
Where as shown in
A going-around component obtained in a state that the speakers 22 and the microphones 21 are suspended in a free space is space propagation sound, and A going-around component obtained in a state that the speakers 22 and the microphones 21 are mounted on the terminal body includes both of space propagation sound and solid propagation sound. Reproduction signals, vibration signals, and vibration going-around data are collected in advance in large numbers. An approximate relationship between the reproduction signal and the vibration going-around component (solid propagation sound) is obtained in advance in the form of a function so that the latter can be calculated from the former.
When the concept of the embodiment is applied to an actual product, a going-around component may be eliminated from a picked-up sound signal by estimating (calculating) a vibration going-around component using a reproduction digital signal and the above approximate function without mounting the vibration sensors 23. With this measure, it is not necessary to mount the vibration sensors 23, whereby a terminal can be produced at a low cost.
A second embodiment will be described below with reference to
The audio processing section 30 has a D/A converter 32 and a feedback canceling section 35 and a feedback cancellation control section 36 which constitute a controller for suppressing noise due to vibration and acceleration. The D/A converter 32 converts a digital audio signal xa[n] as adjusted by the feedback canceling section 35 into an analog signal and outputs the latter to the speaker 22.
The speaker 22, which is a monaural speaker (alternatively, stereo speakers are used), emits a sound (reproduction sound) in the ear where it is inserted. The speaker 22 converts an analog signal which is received from the D/A converter 32 into physical vibration and outputs it as a sound.
The audio processing section 20 also has an A/D converter 33 which is connected to the microphone 21. The microphone 21, which is a monaural microphone (alternatively, stereo microphones are used), picks up a sound that is traveling through the space where the electronic apparatus 110 exists. The microphone 21 converts the picked-up sound into an analog picked-up sound signal and outputs the latter to the A/D converter 33.
The A/D converter 33 converts the analog picked-up sound signal into a digital signal z[n] and outputs the latter to the feedback canceling section 35. The feedback cancellation control section 36 controls the feedback canceling section 35 as the latter generates a noise-suppressed digital audio signal and outputs it to the D/A converter 32.
A pulse wave sensor 34 receives a human pulse wave and outputs a resulting signal to the vital signal clearing processing section 20B. The vital signal clearing processing section 20B performs vital signal clearing processing using an output of the acceleration sensor 16 (to eliminate noise that results from vibration caused by user motion also from the vital signal) and an output of the vibration sensor 23 (to eliminate noise that results from vibration produced by the speaker 22 also from the vital signal), and outputs a resulting signal to a communication section 24B.
As a result, the adaptive feedback canceller 103 can divide an impulse response b̂(n) of a feedback path model for a feedback path (going around) having an impulse response b(n) into an invariable feedback path model having an impulse response f(n) and a variable feedback path model having an impulse response e(n). Therefore, the adaptive feedback canceller 103 can trace a variation of the feedback path (b(n)) using the invariable feedback path model (f(n)) and the variable feedback path (e(n)). A variation in the feedback path (b(n)) is detected on the basis of the acceleration sensor 16, and, if a variation is detected, the degree of update of the filter coefficients of the variable feedback path (e(n)) is increased. Whereas conventionally a digital picked-up sound signal z[n] is used in a feedback canceller as a reference signal, in this embodiment a digital vibration signal xf[n] received from the vibration sensor 23 is also used in the feedback canceller 103 as a reference signal, whereby not only going-around (feedback) sound of space propagation but also going-around (feedback) sound of solid propagation is suppressed.
In this embodiment, the invariable feedback path model may be included in a finite impulse response (FIR) filter or an infinite impulse response (IIR) filter.
The embodiments provide an echo suppressing method which can suppress not only acoustic echo (air propagation sound) that is emitted from a speaker and goes around through an acoustic space and reaches a microphone but also going-around sound (solid propagation sound) from the speaker to the microphones due to apparatus body vibration which cannot be suppressed by any conventional method. As described above, in an environment in which a reproduction signal of a TV receiver, for example, is mixed with music or during a voice call of VoIP, for example, an echo component can be estimated stably and its introduction into a microphone as going-around sound can be suppressed stably. This allows increase of the reproduction sound volume.
(1) Echo due to vibration is eliminated using an output of a vibration sensor as a reference signal.
(2) Echo due to vibration is eliminated by an adaptive filter which uses an output of a vibration sensor as a reference signal.
(3) The echo suppression using the vibration sensor uses an algorithm which takes doubletalk into consideration but not an echo path variation.
(4) Where the acoustic echo canceller using an output signal of a speaker as a reference signal (first echo suppression) is also used, the echo canceller using the vibration sensor (second echo suppression) is disposed downstream of the former.
(5) An acceleration sensor is provided to detect an echo path variation, and the learning of the acoustic echo canceller is controlled according to a detected echo path variation.
(6) Where a vital information sensor is also used, speaker vibration causes noise introduction into the vital information sensor. In view of this, noise is eliminated from a vital signal using a vibration sensor.
The invention is not limited to the above embodiments themselves and may be practiced by variously modifying constituent elements without departing from the spirit and scope of the invention. Various inventive concepts may be conceived by properly combining plural constituent elements disclosed in each embodiment. For example, several ones of the constituent elements of each embodiment may be omitted, and constituent elements of different embodiments may be combined as appropriate.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2012-138184 | Jun 2012 | JP | national |
