Methods and apparatuses for signal analysis
AssigneeHuq Speech Technologies B.V.
InventorsTjeerd Catharinus Andringa, Hendrikus Duifhuis, Pieter Willem J. van Hengel, Michael Gerardus Heemskerk, Maartje Marjolein Nillesen
Life of this patent
Estimating the term…
Abstract
A basilar membrane model is used to receive an input signal including a target signal in step I. With successive further steps the target signal is filtered from the input signal. After the filtering the target signal can be used as an input for further processing, like for example signal recognition of data compression. The target signal can also be applied to a substantially reverse method to be obtain an improved or clean signal.
Claims48
An apparatus for estimating frequency characteristics of an input signal, the apparatus comprising:
- a basilar membrane model device to which the input signal is applied;
- and
- first and second correlator devices, each of said correlator devices connected to a separate segment of the basilar membrane model device and each of said correlator devices having:
- a first input, connected to a respective one of the segments of the basilar membrane model device, for receiving a segment signal from the respective one segment, the segment signal being present for a predetermined period of time;
- and
- a second input, connected to the respective one segment, for receiving the segment signal but shifted by an adjustable time shift;
- and
- wherein the time shift associated with the first correlator device is related to the time shift associated with the second correlator device, the first and second correlator devices collectively providing an output signal which is responsive to the time shifts and to frequencies present in the segment signal applied to each of the correlator devices, and the output signal forms a measure of the frequency characteristics of the input signal.
An apparatus for determining peaks in an input signal, the apparatus comprising:
- a basilar membrane model device to which the input signal is applied;
- first and second integrator devices, each of said integrator devices being connected through a corresponding input thereof to a separate segment of the basilar membrane model device so as to receive a corresponding basilar membrane signal, said each integrator device generating a segment excitation signal from its corresponding basilar membrane signal and transmitting said the segment excitation signal to an output of said each integrator device;
- a peak search device, connected to the outputs of said first and second integrator devices, for determining peaks in the segment excitation signals collectively produced by both integrator devices;
- and
- a correlator device comprising:
- a first input, connected to a respective segment of the basilar membrane model device, for receiving a segment signal, said segment signal being present for a predetermined period of time;
- and
- a second input, connected to said separate segment, for receiving the segment signal but shifted by an adjustable time shift, said correlator device also being communicatively connected to the peak search device such that a segment excitation signal of a segment at a selected position of the basilar membrane device is applied to the correlator device;
- and
- wherein the adjustable time shift, the selected position and a value produced by the respective segment itself vary in time depending on peaks determined by the peak search device.
A method for estimating frequency characteristics of a input signal, the method comprising the steps of:
- receiving the input signal;
- generating a basilar membrane signal, through a basilar membrane model device and in response to the input signal applied to the basilar membrane model, the basilar membrane model device having a number of segments and generating, in response to said basilar membrane signal, an excitation signal from one of said segments;
- generating first and second segment signals of first and second ones of said segments of said basilar membrane signal, respectively;
- generating first and second shifted signals by shifting each of the first and second segment signals, respectively, by an adjustable time shift, the time shift for the first segment signal being related to the time shift for the second segment signal;
- and
- correlating each of the first and second shifted signals with an associated one of the first and second segment signals, respectively, so that a time shift dependent output signal is obtained;
- wherein the output signal is responsive to frequencies present in the excitation signal and forms a measure for the frequency characteristics of the input signal.
A method for determining peaks in a signal, the method comprising the steps of:
- receiving the input signal;
- generating a basilar membrane signal, through a basilar membrane model device and in response to the input signal applied to the basilar membrane model, the basilar membrane model device having a number of segments and generating, in response to said basilar membrane signal, excitation signals from separate ones of said segments;
- determining peaks in the excitation signals;
- and
- for each one of a plurality said segments located at a selected position of the basilar membrane device:
- receiving a segment signal associated with said one segment, said segment signal being present for a predetermined period of time;
- and
- receiving the segment signal shifted over an adjustable time shift;
- wherein the adjustable time shift, selected position and a value produced by the one segment itself vary in time depending on peaks so determined.
Dependencies are inferred from the claim text.