Method and apparatus for suppressing audible noise in speech transmission

• US 6,820,053 B1
• Filed: 10/06/2000
• Issued: 11/16/2004
• Est. Priority Date: 10/06/1999
• Status: Expired due to Term

• Current Assignee: Analog Devices International Unlimited Company (Analog Devices, Inc.)
• Original Assignee: Dietmar Ruwisch
• Inventors: Ruwisch, Dietmar
• Primary Examiner(s): Dorvil, Richemond
• Assistant Examiner(s): Han, Qi
• Application Number:
  US09/680,981
• Time in Patent Office: 1,502 Days
• Field of Search: 704/232, 704/226, 704/202, 706/22, 706/25, 706/31, 381/94.3
• US Class Current: 704/232
• CPC Class Codes: G10L 21/0208 Noise filtering
• CPC Class Codes: G10L 25/30 using neural networks

Abstract

Method of suppressing audible noise in speech transmission by means of a multi-layer self-organizing fed-back neural network comprising a minima detection layer, a reaction layer, a diffusion layer and an integration layer, said layers defining a filter function F(f,T) for noise filtering.

Claims

1. A method of suppressing audible noise during transmission of a speech signal by means of a multi-layer self-organizing feed-back neural network, the method comprising the steps of:

providing a minima detection layer, a reaction layer, and a diffusion layer, and an integration layer,

the minima detection layer for tracking a plurality of minima,

the reaction layer utilizing a non-linear reaction function,

the diffusion layer having only local coupling of neighboring nodes within the diffusion layer, and

the integration layer summing a nodal output of the diffusion layer into a single node without weighting; and

defining a filter function F(f,T) for noise filtering by successively coupling nodes between the minima detection layer, the reaction layer, the diffusion layer, and the integration layer,

wherein f denotes a frequency of a spectral component being analysed at time T.

2. The method as in claim 1, further comprising the step of multiplying an adjustable parameter K with a reaction function in the reaction layer in order to determine an amount of noise suppression of the filter function F(f,T) in its entirety.

3. The method as in claim 1, wherein one node of the integration layer integrates the filter function F(f,T) at a fixed time T over the frequencies f, and wherein a resultant integration signal S(T) so obtained is fed back into the reaction layer.

4. The method as in claim 1, further comprising the step of inputting a spectrum A(f,T) generated by a sampling unit (10) to the minima detection layer, wherein a minima of averaged amplitudes of spectral components A(f,T) averaged over a time corresponding to m frames of an input signal are detected within a given time interval of a length that corresponds to 1 frames of the input signal.

5. The method as in claim 1, further comprising the step of:

using a neural network to generate the filter function F(f,T) from a spectrum A(f,T) being derived by Fourier transformation from a frame of an input signal x(t);

spectrum A(f,T), and the filter function F(f,T) being multiplied to generate a noise-reduced spectrum B(f,T) that, by application of an inverse Fourier transformation in a synthesis unit (12), generates a noise reduced speech signal y(t),

wherein one node of the minima detection layer operates independently from other nodes of the minima detection layer to process a single signal component of the frequency f, and

wherein t denotes the time of handling a sample of the signals x and/or y.

6. The method as in claim 1, further comprising the step of evaluating spectral properties of speech signals in the diffusion layer, the nodes of said diffusion layer effecting frequency component coupling in a manner of diffusion in a frequency domain, with a diffusion constant D>

0.

7. The method as in claim 1, further comprising the step of multiplying all frequency components of filter function F(f,T) at time T with corresponding amplitudes A(f,T), wherein the integration layer effects integration over frequency components of the filter function F(f,T) to produce an integration signal S(T) to be fed back into the reaction layer.

8. The method as in claim 1,

wherein signal components of the speech speech are modulated within modulation frequencies between 0.6 Hz and 6 Hz, an attenuation is less than 3 dB for all values of control parameter K in order to pass the filter function F(f,T) in an optimum manner, the modulation frequencies between 0.6 Hz and 6 Hz corresponding to modulation of human speech, and

wherein the signal components outside of the range of 0.6 Hz to 6 Hz are identified as noise, and are more strongly attenuated based on a value of an adjustable parameter K.

9. An apparatus for audible noise suppression during transmission of a speech signal with a neural network comprising:

a minima detection layer, a reaction layer, a diffusion layer, and an integration layer;

the minima detection layer for tracking a plurality of minima,

the reaction layer utilizing a non-linear reaction function,

the diffusion layer having only local coupling of neighboring nodes within the diffusion layer, and

the integration layer for summing a nodal output of the diffusion layer into a single node without weighting; and

a filter function F(f,T) for noise filtering,

wherein frequency components of a spectrum differ by frequency f and correspond to unique nodes for each of the layers of the network, except for the integration layer, and

wherein each node of the minima detection layer derives a value M(f,T) for the frequency component f at time T, where M(f,T) is obtained by time-averaging an amplitude A(f,T) over a time interval of a length of m frames and a minimum detection of said average within a time interval of the length of 1 frames, with 1>

m.

10. The apparatus as in claim 9, wherein each node of the reaction layer which uses a reaction function r[S(T−

1)] to determine relative spectrum R(f,T) from integration signal S(T−

1), and a freely selectable parameter K, sets an the noise suppression, and from A(f,T) and M(f,T), with relative spectrum R(f,T) having a range of values between zero and one, a formula for determination of R(f,T) being R(f,T)=1−

M(f,T)r[S(T−

1)]K/A(f,T) with the reaction function r[S(T−

1)].

11. The apparatus as in claim 10,

wherein a range of values of the reaction function is limited to an interval [r1, r2], by a reaction function reading r(S)=(r2− r1)exp(S)+r1,

wherein r1 and r2 are arbitrary numbers, and r1<

r2, and

wherein the range of values of the resultant relative spectrum R(f,T) is limited to the interval [0, 1] by setting R(f,T)=1 in case R(f,T)>

1 and setting R(f,T)=0 in case R(f,T)<

0.

12. The apparatus as in claim 10, wherein the nodes of the reaction layer have input thereto an integration signal S(T−

1) from a preceding frame (time T−

1), and are computed in the integration layer and are fed back into the reaction layer.

13. The apparatus as in claim 9, wherein attenuation of the speech signal for all indicated values of control parameter K is lower than 3 dB when speech signals are modulated within modulation frequencies between 0.6 Hz and 6 Hz.