A periodic signal detector is disclosed which may be incorporated to an antage in a voice actuated communication system. It comprises a series connected low pass filter, limiter, and bistable multivibrator, the outputs of the latter of which are respectively connected to the inputs of a pair of ramp generators. A differential comparator detects the difference in the output voltage from said pair of ramp generators and uses it to timely trigger a first monostable multivibrator. A shift register and AND gate extracts a desired program of successive binary signals from the output signal from said first monostable multivibrator. A second monostable multivibrator, triggered by one of the outputs of said bistable multivibrator, in turn, resets said ramp generators and timely shifts said shift register. The output of said AND gate is used to key a transmitter or some other utilization apparatus.
A novel apparatus and method for electronically determining whether audio signals which are present in an electrical audio device (for example, a telephone) are voice signals or, on the other hand, some type of supervisory or system-generated signals. Audio signal lines are connected to an audio detector and a digitizer. The audio detector provides an output which indicates the presence of an audio signal. The digitizer transforms any audio signal into a series of high and low level signals in some manner. The outputs from the audio detector and the digitizer are then provided to a central processing unit which is used to analyze the data. If an audio signal is detected and the digitized audio signal is determined to have a regular periodic pattern, it is presumed that the audio signal is some type of system-generated supervisory tone. If, on the other hand, the digitized audio signal does not have a regular periodic pattern, it is presumed that the audio signal is a voice signal. Accordingly, upon detecting an irregular audio signal, the central processing unit produces an output signal which can be detected by associated circuitry as indicating that voice signals are present.
Speech or voice activity in an audio signal is defected without using a speech coder. Pitch period information and signal energy information are extracted from the audio signal, and a decision regarding the presence or absence of voice is made based on that information.
A waveform-based technique for generating periodicity information from an input signal includes generating a pre-processed signal by applying low pass and non-linear filtering to the input signal, wherein the pre-processed signal has highlighted speech pitch tracks. An adaptive threshold algorithm is applied to the pre-processed signal to generate a detection signal having waveform segments whose peaks are separated by a pitch period of the input signal. A period between peaks in the detection signal is determined that indicates the periodicity information. Information about the period between the peaks in the detection signal is then used to adapt a scaling value to be used by the adaptive threshold algorithm in a subsequent step. The periodicity information may be utilized in a voice activity detector in a telephonic communications system.
