README.md

gender_classification

This repository contains the codes for SMAI course project for the year 2017

packages and installations

In order to run this code you must have anacoda installed in your system. After you are done installing conda just type in the below ommand in the terminal

conda create --name newenv --file requirements.txt

To run the code

python real_time_classify.py -c configs/faces.json -o results/ -d facescrup 

Parameters to be set

1. -c path of configuration file
2. -o path of output directory
3. -d data set to be used

Methodology

We compute eigen face for each face image and use SVM classifier with an rbf kernel to do a binary classifucation between fale and female face images.

The below image shows the genarated eigen faces for the Training dataset

Results

Real time gender classification

To run the BoW code

python test_bow.py configs/faces.json

To run logistic regression

python test_lr.py configs/faces.json