EmoDetect

A project that enables computers to identify human emotion.

All files copyright 2012 Rishabh Animesh, Skand Hurkat, Abhinandan Majumdar, Aayush Saxena.

Licence

EmoDetect is free software: you can redistribute it and/or modify it under the terms of the GNU General Public Licence as published by the Free Software Foundation, either version 3 of the software, or (at your option) any later version.

EmoDetect is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public Licence for more details.

You should have received a copy of the GNU General Public Licence along with EmoDetect. If not, see http://www.gnu.org/licenses

About

We've trained classifiers on over 10,000 images from the MUG Facial Expression Database; the trained classifiers are available in the data/xml folder.

Files

• demo : Binary for Project Demonstration
• extractGaborMain : Extracts Gabor features and dumps them to a file
• extractPHoGMain : Extracts HoG features and dumps them to a file
• extractHaarMain : Extracts Haar features and dumps them to a file
• extractMomentsMain : Extracts Mom features and dumps them to a file
• readImage : Read image from a file and display it (Hello World equivalent for OpenCV)
• test : Test a specific extractor-learner combination on a set of test images
• train : Train from a database of images using a specifi extractor-learner combination and save the classifier to a file
• trainAndCrossValidateRT : Train and Cross Validate for Random Trees (RT) and produce the optimum max_depth
• trainAndCrossValidateSVM: Train and Cross Validate for Support Vector Machines (SVM) and produce the optimum max_depth
• trainAndTestAllAlgos : Train and Test without cross validation using the trained classifer (default)
• trainAndValidate : Train using a saved classifier and measure the test error on a test set
• validateHuman : Executable shows 50 random images and asks the user to guess their right emotion

Run any binary with --help option to see the command line interface for the binary

Compiling

• Package OpenCV (>2.4) installed with the WITH_QT flag set
• CMAKE (>2.6) is required for compilation

CMake will generate the build files for your favourite environment. The code works on Linux, Windows and Mac.