Noise cancelling microphone apparatus
Example embodiments include a method of reducing noise include forming a main signal and one or more reference signals at a beam-former based on at least two received audio signals, detecting voice activity at a voice activity detector, where the voice activity detector receives the main and reference signals and outputting a desired voice activity signal, adaptively cancelling noise at an adaptive noise canceller, where the adaptive noise canceller receives the main, reference, and desired voice activity signals and outputs an adaptive noise cancellation signal, and reducing noise at a noise reducer receiving the desired voice activity and adaptive noise cancellation signals and outputting a desired speech signal.
1. A noise cancelling digital signal processor (DSP), comprising:
a) a beam-forming module configured to output a main signal and one or more reference signals based on audio signals received from at least two microphones, wherein the beam-forming module includes one or more de-emphasis filters, each of which comprises at least one of a low-pass filter and a frequency response matching filter, each of the one or more de-emphasis filters arranged to filter at least one of the main and reference signals;
b) a voice activity detection module configured to receive the main signal and reference signals and to output a desired voice activity signal, and whereby a first one of the one or more de-emphasis filters is arranged to filter the main signal, and a second one of the one or more de-emphasis filters is arranged to filter the reference signal, prior to reception of the main signal and the reference signal by the voice activity detection module;
c) an adaptive noise cancellation module configured to receive the main signal, reference signals and desired voice activity signal and output an adaptive noise cancelled signal; and
d) a single channel noise reduction module configured to receive the desired voice activity signal and adaptive noise cancellation signal and output a desired speech signal.
2. The noise cancelling DSP of claim 1 , wherein the desired voice activity signal is further configured to control the adaptive noise cancellation module and the noise reduction module.
3. The noise cancelling DSP of claim 1 , wherein the voice activity detection module further includes:
a) one or more short-time detectors, configured to detect a short-time power of each of the received main and reference signals, respectively;
b) one or more log scalers, configured to convert the detected short-time power of each short-time detector, respectively; and
c) one or more combiners, configured to receive the log scaled short-time power detections of the main signal and one of the reference signals and produce a voice activity difference signal based on said detections.
4. The noise cancelling DSP of claim 3 , wherein the voice activity difference signal is further communicatively coupled to a signal channel voice activity detection module outputting the desired voice activity signal.
5. The noise cancelling DSP of claim 4 , wherein the voice activity detection module further includes an OR-gate or an AND-gate, based on microphone configuration, arranged to receive multiple desired voice activity signals and output one of the multiple desired voice activity signals.
6. The noise cancelling DSP of claim 3 , wherein the short-time detector is a root-mean-square detector, power detector, or energy detector.
7. The noise cancelling DSP of claim 1 , wherein the low-pass filter adjusts at least one of the main signal and the reference signal to accommodate a close talk range of the at least two microphones.
8. The noise cancelling DSP of claim 7 , wherein the de-emphasis filter has a gradual slope roll-off starting at a frequency approximately between 1 kHz and 4 kHz and continuing to approximately a Nyquist frequency.