Recognition - Multimedia Signal Processing - Lecture Slides, Slides of Electronics engineering

Download Recognition - Multimedia Signal Processing - Lecture Slides and more Slides Electronics engineering in PDF only on Docsity!

CONTENT BASED FACE

RECOGNITION

Introduction

Problem Statement :

• Given an image, to identify it as a face and/or extract face images from it.
• To retrieve the similar images (based on a heuristic) from the given database of face images.

 Faces are complex, multidimensional and meaningful visual stimuli.

 Face Recognition is difficult.

 Face Images are similar in overall configuration.

Difference From Image Recognition

Approach

• Similar to Content Based Image Retrieval (CBIR).
• Neural Networks and Self Organizing Maps (SOMs).
• Principal Component Analysis (PCA).
• Relevance feed back.

Face Recognition Using

Eigenfaces

 Face Images are projected into a feature space (“Face Space”) that best encodes the variation among known face images.

 The face space is defined by the “eigenfaces”, which are the eigenvectors of the set of faces.

Eigen Space and Eigen Faces

PCA

Main assumption of PCA approach:

• Face space forms a cluster in image space.
• PCA gives suitable representation.

Eigenfaces (1)

• Calculation of Eigenfaces (1) Calculate average face : v. (2) Collect difference between training images and average face in matrix A (M by N), where M is the number of pixels and N is the number of images.

(3) The eigenvectors of covariance matrix C (M by M) give the eigenfaces.

• M is usually big, so this process would be time consuming.

What to do?

C = AA^ T

Eigenfaces (3)

• Representation of Face Images using Eigenfaces
• The training face images and new face images can be represented as linear combination of the eigenfaces.
• When we have a face image u :

Since the eigenvectors are orthogonal :

= (^) ∑ i

u ai φ i

i

T ai = u φ Docsity.com

Eigenfaces (4)

• Experiment and Results Data used here are from the ORL database of faces. Facial images of 16 persons each with 10 views are used. - Training set contains 16×7 images.
  • Test set contains 16×3 images.

First three eigenfaces :

Neural Networks

and

TS-SOM

What are Neural Networks?

• Individual units to simulate Neurons
• Parallel Processing
• Many inputs and single output
• Organization/structure of the TLU’s is important

Training of SOM

• Randomly initialized
• Selection based on some query parameter
• On selection a node and its neighbors are modified
• Degree of modification reduces with each iteration

Example of a two-dimensional TS-SOM structure of 3 levels