Eigenfaces: Facial Feature Extraction

Configuration

Camera Input

Captured images: 0 (need at least 2)

Captured Image

Eigenfaces

These are the principal components extracted from the captured faces:

Adjust Eigenface Weights

Modify the sliders to see how different eigenfaces contribute to face reconstruction:

How Eigenfaces Work

Principal Component Analysis

Eigenfaces use PCA to identify patterns in facial images by:

  1. Computing the average face (mean)
  2. Finding differences between each face and the average
  3. Calculating the principal components of variation
  4. Using these components as a basis for face recognition
Face Recognition Process

To recognize a face, the system:

  1. Projects the face onto the "eigenspace"
  2. Computes a set of weights (the face signature)
  3. Compares this signature to known faces
  4. Identifies the closest match