Skip to content

Eigenfaces and Fisherfaces for Face Recognition. With implementation of algorithms as PCA, KNN, Fisher Discriminant Analysis.

License

Notifications You must be signed in to change notification settings

vvarodi/face_recognition

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Eigenfaces and Fisherfaces for face recognition

Table of Contents:

  1. Introduction
  2. Content
  3. Eigenfaces and Fisherfaces obtained

1. Introduction

For this project, an image dataset was provided in order to perform a classification task by building a statistical model. This face recognition model must be able to tell us if a given image belongs to this dataset or not. If the face image is a member of the dataset, the model must return the corresponding label associated with that face. On the other hand, a 0 is returned when the image does not belong to any of our individuals, we have implemented thresholds to recognize this impostor. The Training dataset is a subsample of 25 individuals of the original dataset: faces94.

The statistical model was build using different techniques:

  • Principal Component Analysis (PCA)
  • K-Nearest Neighbours (KNN)
  • Fisher Discriminant Analysis (FDA)

Combining algorithms, two different models were obtained:

  1. PCA + KNN
  2. PCA + FDA + KNN

In both models, cross validation and parameters optimization were performed.

To see the results, conclusions and procedure to understand how it was done, you can see the report.

2. Content

* [Images]    Images used in the report / README
* [Training]  Training Dataset provided
* [RData]     Utils needed in model.R, load before using face_recognition() function
* report.pdf  Results, conclusions and procedure   

* training.R  Source code
* model.R     Final model, face_recognition() final function

3. Eigenfaces and Fisherfaces obtained

TO DO:

  • Work in the documentation
  • Provide the Training images
  • Provide the real full dataset, and images to test our faces and another impostors