Contents


Detailed Discussion

 

Title Page

Abstract

Introduction

Feature Selection

Optimality Criterion

Approach Overview

Overview of Results

 

Case Study: Classification of Race

Race Training Data

Neural Networks for Classifying Race

Neural Network Results

Analysis of the Neural Network Results

Multi-Class GLD for Classifying Race

Forming a Multi-Class GLD Classifier

Multi-Class GLD Results

Back to the Drawing Board...

The Idea: Synthetic Faces

Synthetic Faces via Averaging

Race Classification with an Augmented Training Set

Sorting Out the Confusion...

New Feature: Average Pixel Intensity

Classification of Race: Summary

 

Other Results with Synthetic Faces

 

Summary and Conclusions

 

 

 

 

Return to Start
Detail