Digital Signal Processing

- Project -

Slot: G1+TG1
Professor: Malaya Kumar Hota
 Hearing Aid for Impaired People Using “Sound
Manipulation Algorithm”
By

A.L.P
LIKITH S.V
ABHIRAM
17BEC0386
17BEC0578
 Abstract

•Traditional analog hearing aids are like a simple radio. They can be tuned and adjusted for volume, bass and
treble. But hearing loss is not just a technical loss of volume. Rather, hearing deficiency can increase sensitivity
and reduce tolerance to certain sounds while diminishing sensitivity to others. For instance, digital technology can
tell the difference between speech and background noise, allowing one in while filtering out the other.
•Approximately 10% of the world's population suffers from some type of hearing loss, yet only a small percentage
of this statistic use a hearing aid. The stigma associated with wearing a hearing aid, customer dissatisfaction with
hearing aid performance, and the cost associated with a highperformance solution are all causes of low market
penetration.
• Using digital signal processing, digital hearing aid now offers what the analog hearing aid cannot offer. It
proposes the possibility of performing signal-to noise enhancement, flexible gain-processing, digital feedback
reduction, etc. In this paper, the simulation of simple digital hearing aid was developed using MATLAB
programming language.
• The implementation of this configurable digital hearing aid (DHA) system includes the noise reduction filter,
frequency shaper function, and amplitude compression function. This digital hearing aid system is designed to
adapt for mild and moderate hearing loss patient since different gain can be set to map different levels of hearing
loss
 INTRODUCTION

Hearing Aids systems are one of the most important issues for human being. They are a small electronic instrument which
makes sound louder and makes speech easier to hear and understand. The hearing aid is designed to pick up sound waves
with a tiny microphone, change weaker sounds into louder sounds and send them to the ear through a tiny speaker. With
the microchips available today, hearing aids have gotten smaller and smaller and have significantly improved quality.
Roughly 10% of the world population bears from some hearing loss. However, only a portion uses hearing aid. This is due
several factors which include the stigma associated with wearing a hearing aid, customer dissatisfaction with the devices not
meeting their expectations, and the cost associated with the new digital versions of hearing aids. Hearing loss is typically
measured as the shift in auditory threshold relative to that of a normal ear for detection of a pure tone. Therefore, there are
many types of hearing aids with a wide range of functions and features to address individual needs. Table 1 shows the
classification of degrees of Hearing Loss. A hearing aid is an electronic device that makes sounds louder and can help to
offset hearing loss. The aim of the hearing aid is to amplify sound signals in such a way that they become audible for the
hearing partially impaired person. Classification of Hearing Loss Hearing level Normal hearing -10 dB – 26 dB mild hearing
loss 27 dB - 40 dB Moderate hearing loss 40 dB - 70 dB severe hearing loss 70 dB - 90 dB Profound hearing loss greater than
90 dB Different degree of Hearing Loss
Block Diagram
A hearing aid is an electronic device that makes sounds louder and can help to offset hearing loss. The aim of the hearing aid
is to amplify sound signals in such a way that they become audible for the hearing partially impaired person .
Different degree of Hearing Loss:

Classification of Hearing Hearing level

Loss

Normal hearing -10 dB – 26 dB

mild hearing loss 27 dB - 40 dB

Moderate hearing loss 40 dB - 70 dB

severe hearing loss 70 dB - 90 dB

Profound hearing loss greater than 90 dB

 Methodology

The input speech signal takes the form of human voice. The input speech signal will pass through
several functions i.e noise addition, noise reduction filter, frequency shaper and amplitude compression
before producing an adjusted output speech signal which is audible to the partially hearing-impaired
person.

NOISE ADDITION: Since the input speech signal for this system is a clean signal, some noise is added
in order to simulate a real situation. In this system, the Adaptive White Gaussian Noise (AWGN) and
random noise are added to the input speech signal by using MATLAB function. Noise (AWGN) has a
continuous and uniform frequency spectrum over a specified frequency band and has equal power per
Hertz of this band. It consists of all frequencies at equal intensity and has a normal (Gaussian)
probability density function
NOISE REDUCTION FILTER: A major anxiety for the people with hearing loss is the capability of
hearing aid to differentiate intended speech signal in a noisy environment. Hence, to eliminate the
noise, a reduction filter function is used in this design. To suppress the noise in the signal, the wavelet
filter function is used.
FREQUENCY SHAPER:
One major complaint of hearing aid users is that the hearing aid amplifies all signals
rather than the significant signal that they desire to hear. Most hearing impaired has
difficulties to hear high frequency signal. Therefore, the frequency shaper is
designed to correct for loss of hearing at certain frequencies. It applies high gain for
higher frequencies and vice versa.

AMPLITUDE COMPRESSION :
Fundamentally, amplitude compression function is the task of controlling the overall
gain of a speech amplification system. Amplitude compression will ensure that the
amplified signal will not exceed saturation power. Saturation power is where the
sound signal begins to become uncomfortable..
 Matlab Codes
DENOISING A SIGNAL :
function y = denoiseEm(x);
[thr,sorh,keepapp]=ddencmp( 'den' , 'wv' ,x);
y=wdencmp( 'gbl' ,x, 'db3' ,2,thr,sorh,keepapp);
subplot(2,1,1) plot(x);
subplot(2,1,2) plot(y);

FREQUENCY SHAPING:
function y = applySkiSlope(x,g,transitionV,fs);
first = transitionV(1);
second = transitionV(2);
third = transitionV(3);
fourth = transitionV(4);
x_length = length(x);
n = nextpow2(x_length);
N = 2^n; T = 1/fs;
X = fft(x,N); gain = zeros(N,1);
% Sets the gain for the first stage of frequencies:
firstC = (.3*(g-1))/first;
k=0;
while(k/N <= first/fs) gain(k+1) = firstC*k/(N*T) + 1;
gain(N-k) = gain(k+1);
k=k+1;
end;

% Sets the gain for the second stage of frequencies:

secondC = firstC*first +1; secondC2 = (second-first)/5;
while(k/N <= second/fs) gain(k+1) = 1 + (secondC-1)*exp(-((k/(N*T))-first)/secondC2);
gain(N-k) = gain(k+1);
k=k+1;
end;

% Sets the gain for the third stage of frequencies:

thirdC = 1 + (secondC-1)*exp(-second/secondC2);
thirdC2 = (third-second)/5;
while(k/N <= third/fs) gain(k+1) = g + (thirdC-g)*exp(-((k/(N*T)-second))/thirdC2);
gain(N-k) = gain(k+1);
k=k+1;
end;
% Sets the gain for the fourth stage of frequencies
while(k/N <= fourth/fs) gain(k+1) = g;
gain(N-k) = gain(k+1);
k=k+1;
end;

% Sets the gain for the fifth stage of frequencies

fifthC = g;
fifthC2 = (fs/2-fourth)/5;
while(k/N <= .5) gain(k+1) = 1 + (fifthC-1)*exp(-((k/(N*T))-fourth)/fifthC2);
gain(N-k) = gain(k+1);
k=k+1;
end;
k_v = (0:N-1)/N;
plot(k_v,gain); %entire filter transfer function
figure; %non-redundant filter transfer function
k_v = k_v*fs; k_v = k_v(1:N/2+1);
plot(k_v,gain(1:N/2+1));
title('Frequency Shaper Transfer Function');
xlabel('Frequency (Hertz)');
ylabel('Gain');
Y = X+gain;
y = real(ifft(Y,N));
y = y(1:x_length);
t=[0:1/fs:(x_length-1)/fs];
figure;
plot(t,y,'r'); %hold; figure; plot(t,x);
AMPLITUDE COMPRESSION:
function y = powerCompress(input, Psat,Fs);
x=input;
len=Fs*0.1;
iter=floor(length(x)/len);
Plow=0.008;
for rg=0:1:iter;
start=rg*len+1;
en= rg*len+len;
if rg*len+len>length(x) en=length(x);
end clear signal X X_pow Y_pow Y y z X_phase;
signal=x(start:en);
n = nextpow2(len);
N = 2^n; X = fft(signal,N);
X_phase=angle(X); % Save the old phase information
X_pow = abs(X)/N;
Y_pow = X_pow;
Y=zeros(N,1);
for k=0:N/2 if Y_pow(k+1)Psat % Clip amplitudes higher than Psat
Y_pow(k+1)=Psat;
Y_pow(N-k)=Psat;
end;
Y(k+1) = Y_pow(k+1)*(cos(X_phase(k+1))+i*sin(X_phase(k+1)));
Y(N-k) = Y_pow(N-k)*(cos(X_phase(N-k))+i*sin(X_phase(N-k)));
end;
y = real(ifft(Y,N));
z = y(1:en-start+1); sig_out(start:en)=z; end; y = sig_out*2000
MAIN CODE:

function y = hearingAidF(input,g,Psat,transitionV,newfile);
[x,fs] = audioread(input);
xc = denoiseEm(x); % denoising filter
xf = applySkiSlope(xc,g,transitionV,fs); % frequency shaping filter
y = powerCompress(xf, Psat,fs); % amplitude shaping filter
x_length = length(x);
t=[0:1/fs:(x_length-1)/fs];
%sound(y,fs);

% plots for the input and output signals:

figure; subplot(2,1,1);
plot(t,x,'b');
axis tight;
xlabel('Time (sec)');
ylabel('Relative Magnitude');
title('Time Profile for Data in Signal 2');
subplot(2,1,2);
plot(t,y,'r');
axis tight;
xlabel('Time (sec)');
ylabel('Relative Magnitude');
title('Time Profile for Data in Adjusted Signal 2');
figure;
subplot(2,1,1);
specgram(x);
title('Spectrogram of Original Signal 2');

subplot(2,1,2);
specgram(y);
title('Spectrogram of Adjusted Signal 2');

%soundsc(input, fs);
sound(y,fs);
%audiowrite(y,fs,nbits,'linear',newfile);
audiowrite('temp_file.wav',y,fs);
OUTPUT
CONCLUSION

The newer digital aids offer more ability to fine- tune the sound without distorting
the quality and help the listener. In this digital hearing aids system implementation
using MATLAB, sound processing is digitalized. Thus, it is possible to refine the sound
signal, for instance by reducing noise and improving speech signals. In addition, by
using digital technology, the amplification can be done only at the frequencies that
the user needs to amplify. This will eliminate the problem with conventional amplifier
which amplified the whole signal including noise. In general, digital hearing aids,
when the incoming signals are converted to digital signals. This digitalization makes it
possible to precisely analyse & filter the signals. The signals can be processed in one
or more frequency channels. At the end, the digital signal is again converted to its
analog form. The benefits of using digital aids could Improve quality of life by
improving sound quality, Higher listening comfort, better communication in noisy
environment, better speech intelligibility in group conversations and more flexibility
in case of progressive hearing less.
REFERENCES

 IEEE EXPLORER
 Audio, Speech and Language Processing, IEEE Transactions.
 International Journal of Computing & Information Sciences.
 IEEE TRANSACTIONS ON NEURAL SYSTEMS AND
REHABILATION ENGINEERING ,VOL26,NO3.MARCH 2018
THANK YOU