README file

Project : Face Expression Recognition with Multiclass SVM 

Please note that in order to run these files successfully make sure your Matlab has the Computer Vision Toolbox and the Statistics and Machine Learning Toolbox

Also please note that while running the command the current working directory should be changed 
to the directory containing all the extracted Matlab files 
The Face Expression Recognition  was built using Matlab2015b
The dataset used for this project is the Cohn- Kanade images which can be obtained from the following link
http://www.consortium.ri.cmu.edu/ckagree/



DatasetLoad -   This program returns all the true emotion labels of the dataset and saves them in EmotionLabels.mat
eigenfaceFeatures -         This function extracts features corresponding to the eigen faces from the images
EmotionLabels.mat -     This data contains the true class of emotion labels from the ground truth
eval_mcr -                This function evaluates the Miss-Classification Rate
Face_imgFeatures -                This program detects the region of the face in the image and extracts features corresponding to the boundary pixels from Morphological operators, Gabor features, Haar wavelet features  and Landmark points 
featureCombination -              This program combines the various features extracted into different combinations
featureNormalize -                This function is required as a preprocessing step for PCA used in featureCombination.m 
gaborFeatures -               This function is used in Face_imgFeatures.m program to extract Gabor Features from an image 
gaborFilterBank -             This function is used in Face_imgFeatures.m program to extract Gabor Features from an image   
MulitclassSVMClassificationEmotion -           This program performs Mulitclass SVM classififcation with the feature combinations generated by  feautreCombination.m using the one vs. rest method
projectData -                This function is used for PCA in feautreCombination.m 
regiongrow -                 This function is used in Face_imgFeatures.m program to extract Boundary pixels as features using Morphological operations
tofloat -                 This function is used in Face_imgFeatures.m program to extract Boundary pixels as features using Morphological operations