Signal Processing Stack Exchange is a question and answer site for practitioners of the art and science of signal, image and video processing. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I have a large face database, and I want to search through it. For example, if I input a new face image, the system would return the best N similarities.

I know that PCA, LDA, and Procrustes analysis are used for this, but it seems that I need to project all this large database into PCA space, and I don't know any library that can handle that.

Are there some tools, software, or a library for this?

share|improve this question
up vote 1 down vote accepted

If you can transform your problem into the following "given a set of vectors find the N ones with the smallest euclidian distance to the query", there are well-known methods (Ball-trees, Locality-sensitive hashing...) for solving it on large datasets. With this approach, PCA / LDA is thus just a pre-processing step for transforming the set of features produced by your feature extractor into a lower-dimensionality feature vector.

share|improve this answer
I want to use model like ASM(active shape model) for face recognition. For example I mark by the hands points on 1st image and 2nd and then I want to compare them. What metric I should choose? My 1st assumption was to use simple euclidean distanse for 2D points, but before I need to "align" models for example calculate сetres of gravity and align, and then apply scale changes,also I can align by the axes of the ellipses, but I think it doesn't give good align results, I think I need some non linear cost function to "morph" from 1 model to another. – mrgloom Nov 12 '12 at 10:32

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.