Discover MakerZone

MATLAB and Simulink resources for Arduino, LEGO, and Raspberry Pi

Learn more

Discover what MATLAB® can do for your career.

Opportunities for recent engineering grads.

Apply Today

Thread Subject:
ordering eigenfaces

Subject: ordering eigenfaces

From: shahnaz fatima

Date: 29 Mar, 2011 17:13:04

Message: 1 of 2

i have a database of 400 faces images .i represent each facial image as a vector of 10304x1.

i subtract the mean from all columns corresponding to each image and merge them to form a matrix 10304x400.

then i find L and eigen vectors and eigen values of L
 L = A'*A; % L is the surrogate of covariance matrix C=A*A'.
[V D] = eig(L)

then i ty to limit the number of eigenvectors using some threshold.

and then find the eigenfaces as eigenfaces= A*V ;

>>my first doubt

i am getting only 399 eigenfaces and not 400.

why 1 lesser????
>> my second doubt,
i=1:400 and
eigenfaces(:,i)

in what order are the eigenfaces?

i mean how do i get the 10 faces of the first person and so on.

please explain

Subject: ordering eigenfaces

From: Florin Neacsu

Date: 29 Mar, 2011 17:44:05

Message: 2 of 2

"shahnaz fatima" <shahnaz1981fat@gmail.com> wrote in message <imt3v0$nh3$1@fred.mathworks.com>...
> i have a database of 400 faces images .i represent each facial image as a vector of 10304x1.
>
> i subtract the mean from all columns corresponding to each image and merge them to form a matrix 10304x400.
>
> then i find L and eigen vectors and eigen values of L
> L = A'*A; % L is the surrogate of covariance matrix C=A*A'.
> [V D] = eig(L)
>
> then i ty to limit the number of eigenvectors using some threshold.
>
> and then find the eigenfaces as eigenfaces= A*V ;
>
> >>my first doubt
>
> i am getting only 399 eigenfaces and not 400.
>
> why 1 lesser????
> >> my second doubt,
> i=1:400 and
> eigenfaces(:,i)
>
> in what order are the eigenfaces?
>
> i mean how do i get the 10 faces of the first person and so on.
>
> please explain

Hi,

Are you doing a Principal component analysis ?
 And does it really shock you that you don't get 400 eigenvectors ? For a n*n how many eigenvectors can you expect ? Is it even possible to guess ? Besides if you are doing a PCA, would you even want to get 400 ?

Regards,
Florin

Tags for this Thread

What are tags?

A tag is like a keyword or category label associated with each thread. Tags make it easier for you to find threads of interest.

Anyone can tag a thread. Tags are public and visible to everyone.

Contact us