MICROPHONE ARRAY SYSTEM
BACKGROUND AND PRIOR ART
1. Field of the Invention
This invention relates to a microphone array system.
2. Description of the Related Art
Normally a microphone is used to acquire a sound. However, the surrounding environment can largely affect the sound acquisition. Background noises and reverberation of the speaker's voice can degrade the acquisition performance, particularly for a distant-talking. For this reason, a short range microphone designed for use near the mouth of the speaker such as a headset microphone or hand microphone has been employed. However, it is uncomfortable to wear the headset microphone on head, while the hand microphone limits the freedom of the speaker as it occupies his/her hands. In addition, it is also inconvenient for multiple speakers for a conference. There has been a demand for a sound input scheme that can allow more freedom and convenience to the speakers.
A known microphone array system has been studied as a potential candidate for sound input scheme that can solve the conventionally encountered inconvenience described above. The microphone array system consists of a set of a plurality of microphones which are arranged at spatially different positions, where the desired speaker can be localized and tracked, and noises can be reduced by synthetic processing of outputs of these microphones.
The algorithms for reducing noises by microphone array include conventional delay and sum beamforming, broadband adaptive beamforming and match-filtering etc. However, the delay and sum bearrm^rming can provide low performance, broadband adaptive
beamforming needs a lot of computations and is sensitive to the microphone dismatch and microphone position errors, while match-filtering needs to know the information of the transfer functions between the sound source and microphones in prior.
SUMMARY OF THE INVENTION
It is accordingly an object of the present invention to provide a microphone array system to overcome at least a part of said inconveniences. It is another object of the present invention to provide a microphone array system to be able to achieve a good spatial selectivity and provide a high signal-to-noise ratio.
This object is solved by providing a first and a second set of microphones for receiving first and second sets of acoustical signals, an adaptive signal processor, a beamforming circuitry and a steering means. Said adaptive signal processor receives a first set of digital signals resulting from said first set of acoustical signals and outputs both, a localization signal and a processed signal. Said beamforming circuitry receives said localization signal and a second set of digital signals resulting from said second set of acoustical signals and outputs a result signal. Said steering means is controlled by said localization signal and adjusts said second set of microphones into a direction where said first set of acoustical signals comes from.
More specifically, according to an embodiment of the present invention, there is provided a steering microphone array system which comprises a first array composed of omni- directional microphones and a second array composed of directional microphones, preamplifiers, A/D converters, a D/A converter, a video camera (optionally), a small motor to drive the rotation of the fixed directional microphone array, an adaptive signal processor and beamforming circuitry. Said second array composed of directional microphones is of a fixed beam to its direct look direction or such and it is steered (for example: by rotating) by a steering means like a motor and braces according to the localization obtained from said first array composed of omni-directional microphones.
Other objects, features and advantages according to the present invention will be presented in the following detailed description of the illustrated embodiments when read in conjunction with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 illustrates a block diagram of the microphone array system of the present invention;
FIG. 2 illustrates a block diagram of said first array composed of omni-directional microphones;
FIG. 3 illustrates a block diagram of said second array composed of directional microphones; FIG. 4 illustrates a structure diagram of the arrangement of a jώ directional microphone and said motor;
FIG. 5 illustrates a structure diagram of another arrangement for said second array.
DETAILED DESCRIPTION OF THE INVETION
FIG. 1 illustrates the block diagram of the microphone array system of the present invention. The system consists of two arrays 1, 2 of microphones. Said first array 1 is composed of m omni-directional microphones OMOl, ..., OMOm and said second array 2 is composed of n directional microphones DM01 , ..., DMOn.
A first set of acoustical signals received by said m omni-directional microphones OMOl, ..., OMOm is sent to a first set of signal converting means comprising a first set of m preamplifiers OP01, ..., OPOm and a first set of m A/D converters OA01,..., OAOm and converting said (pre-amplified) first set of acoustical signals to a first set of digital signals xl(i), ..., xm(i). An adaptive signal processor 4 receives said first set of digital signals xl(i), ..., xm(i) and outputs both, a localization signal od(i) and a processed signal oz(i) at first and second output lines 6 and 7, respectively. Said adaptive signal processor 4
determines, whether a voice signal exists, and estimates in a case, where such an acoustical signal exists, from which direction said acoustical signal comes.
A second set of acoustical signals (most of these signals originate from the same source {speaker} as said first set of acoustical signals) received by said n directional microphones DM01, ..., DMOn is sent to a second set of signal converting means comprising a second set of n preamplifiers DP01, ..., DPOn and a second set of n AID converters DA01, ..., DAOn converting said (preamplified) second set of acoustical signals to a second set of digital signals yl(i) yn(i).
A beamforming circuitry 5 receives said second set of digital signals yl(i), ..., yn(i) and said processed signal oz(i) from said second output line 7, and outputs a result signal zz(i) at an output 8. Said result signal zz(i) is then sent to an output means 11 (for instance: to a connector) through a D/A converter 10.
A motor 3 (being part of a steering means 2a, 3) is connected to said second array 2, for instance via a brace 2a (said brace 2a also being part of said steering means 2a, 3). Said motor 3 is used to steer a rotation of said second array 2 controlled by said localization signal od(i) from said first output line 6. Said localization signal od (i) from said first output line 6 may, additionally, be used to turn a video camera 9 to focus to the desired speaker.
FIG. 2 illustrates a block diagram of said first array 1 composed of said first set of m omni-directional microphones OMOl, ..., OMOm. Said m omni-directional microphones OM01 , ... , OMOm receive said first set of acoustical signals and convert it to a first set of electronic signals. Said first set of electronic signals is amplified by said first set of m preamplifiers OP01, ..., OPOm and then converted to said first set of digital signals xl(i), ..., xm(i) by said first set of m A/D converters OA01, ..., OAOm.
Within said adaptive signal processor 4 an adaptive voice detection circuit 41 receives said first set of digital signals xl(i), ..., xm(i) and detects, whether a (for example:
human's) voice like signal exists or not. The result of said detection is outputted as a detected signal v(i). If said detected signal v(i)=0, no voice signal was detected. If said detected signal v(i)=l, a voice signal was detected.
A time-delay based localization estimation circuit 42 receives said first set of digital signals xl(i) xm(i) and said detected signal v(i), estimates the localization of the desired speaker (i. e.: the direction, where said first set of acoustical signals comes from), creates and outputs said localization signal od(i) at said first output line 6.
The operation of said time-delay based localization estimation circuit 42 is described as follows:
1) If said detected signal v(i)=0, said time-delay based localization estimation circuit 42 does not actualize its status (i) and it outputs a previous status od(i-l) of said localization signal od. 2) If said detected signal v(i)= 1 , said time-delay based localization estimation circuit 42 actualizes its status (i), estimates the localization of the desired speaker (i. e.: the direction or source, where said first set of acoustical signals comes from), based on time delays between different ones of said omni-directional microphones OMOl, ..., OMOm and outputs an actual localization signal od(i) having been actualized compared to its previous status.
A delay and sum beamforming circuit 43 receives said first set of digital signals xl(i), ..., xm(i) and said detected signal v(i), processes a delay and sum beamforming operation and outputs said processed signal oz(i) at said second output line 7.
FIG. 3 illustrates a block diagram of said second array 2 composed of said second set of n directional microphones DM01, ..., DMOn. Said n directional microphones DM01
DMOn receive said second set of acoustical signals and convert it to a second set of electronic signals. Said second set of electronic signals is amplified by said second set of n preamplifiers DP01 , ... , DPOn and then converted to said second set of digital signals yl(i), ..., yn(i) by said second set of n AID converters, DA01 ..., DAOn.
Said processed signal oz(i) from said adaptive signal processor 4, associated with said first array 1 of omni-directional microphones OMOl, ..., OMOm, is delayed at a delay circuit 52 and weighted at a weight circuit 53 to generate a weighted signal ou(i). A fixed look-direction beamforming circuit 51 receives said second set of digital signals yl(i) yn(i) and said weighted signal ou(i) and does beamforming at the look direction (i. e.: the direction or source, where said first set of acoustical signals comes from). Said motor 3 steers said second array 2 into the direction of the desired speaker (i. e., where said first set of acoustical signals comes from) by means of, for example, said braces 2a, controlled by said localization signal od(i) from said first output line 6.
Said fixed look-direction beamforming circuit 5 outputs said result signal zz(i). Said result signal zz(i) goes through said D/A converter 10 and then leaves the microphone array system at said output means 11 (which may be a connector) as a final result signal of the microphone array system.
FIG. 4 illustrates a structure diagram of the arrangement of one of said directional microphones, for instance microphone DMOj, furtheron called as "fh directional microphone", and said motor 3. Said/Λ directional microphone DMOj is of a directivity pattern DM0j_l which has a main beam and two nulls (in Fig. 4 only one of said nulls is labeled by the word "NULL"). Said motor 3 is installed in the direction of one of said nulls of said/'1 directional microphone DMOj in order to reduce the influence due to the rotating noise of said motor 3 into said A directional microphone DMOj. Another way is to put said motor 3 at the null direction of the whole directivity pattern.
FIG. 5 illustrates a structure diagram of another arrangement for said second array 2. In this arrangement, said directional microphones DM01, ..., DMOn are positioned along an inner circuit. This arrangement is more suitable for near field applications such as a small room, car, computer workstation or the like.