error in speech recogniion
2 次查看(过去 30 天)
显示 更早的评论
I am working on speech recognition. I have completed the training session and created the database. Now for the testing session my code is giving the error
??? Index exceeds matrix dimensions.
Error in ==> readnewfile at 28
display(words(index,:))*
The code is as follows
clc
clear all
close all
load ('Database.mat')
% load ('AudioList.mat')
Datasize=size(Dataset);
[s1,fs] = wavread('wh1.wav'); % Reading new file x1.wav
B = [1 -0.95];
s1 = filter(B,1,s1);
[c]=melcepst(s1,fs);
n=corrcoef(c) ;% Finds the correlation coefficients
n=n(:);
n=n';
distance=zeros(1,5);
word1='Transmit';
word2='What ';
word3='Colour ';
word4='Computer';
words=vertcat(word1,word2,word3,word4);
for i=1:Datasize(1)
distance(1,i)=sqrt(sum((Dataset(i,:)-n(1,:)).^2));
end
[mindistance,index]=min(distance);
display('Detected word is ')
display(words(index,:))
Can someone plz guide me?
2 个评论
Geoff Hayes
2016-7-3
aliha - the error seems to suggest that index is greater than the number of rows in your matrix. What are the dimensions for words and for Dataset?
采纳的回答
Walter Roberson
2016-7-3
You initialize
distance=zeros(1,5);
so distance will be at least of length 5 and will start out containing zeros. You then loop over i=1:Datasize(1) setting distance(i) . If Datasize(1) is less than 5 then you would be leaving a 0 untouched at index 5, and it is not possible that any distance is less than 0 so that 0 would be the minimum distance. If Datasize(1) is 5 or higher than you would be removing all of the 0 that you assigned but you run the risk that the minima is from 5 onwards. You use the index of the minima to index the 4 rows of words so if the index came back 5 or more that is going to fail.
3 个评论
Walter Roberson
2016-7-3
Create a box with 5 drawers. Put objects in 4 of the drawers. Which drawer in the box has the fewest number of objects in it? Answer: the one you did not put any objects into.
Just change your
distance=zeros(1,5);
to
distance=zeros(1,4);
and try again.
更多回答(0 个)
另请参阅
类别
在 Help Center 和 File Exchange 中查找有关 Speech Recognition 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!