how to remove silent regions from speech signal dynamically

%program to remove silences from a speech recording %step 1- break the signal into frames of 0.1seconds fs=11250; %fs=22050; frame_duration=0.1; frame_len=frame_duration*fs; N=length(ip); num_frames=floor(N/frame_len); new_sig=zeros(N,1); count=0; %frame1=ip(1:frame_len); %frame2=ip(frame_len+1 :frame_len*2) %frame3=ip(frame_len*2+1 :frame_len*3) for k=1:num_frames %extract a frame of speech frame=ip( (k-1) *frame_len +1 : frame_len*k); %step2 - identify silence by finding frames with max amplitude less than 0.03

max_val=max(frame);

if(max_val>0.03) %this frame is not silent count=count+1;

new_sig( (count-1)*frame_len +1 :frame_len*count)=frame;

end end

here 0.03 is set as threshold instead how can we do it dynamically.

Thank you

Answer 1

[Image Analyst]

Would this "File Exchange Pick of the Week" be of any use to identify and discriminate between silence and valid sounds: http://blogs.mathworks.com/pick/2014/05/23/automatic-activity-detection-using-hilbert-transform/