Automatic Analysis of Human Behavior
-------------------------------------------------------------------------------
    Coded by Md. Iftekhar Tanveer ([email protected])
    Rochester Human-Computer Interaction (ROCHCI)
    University of Rochester
-------------------------------------------------------------------------------
OS Compatibility:
================
Code tested in OS X 10, Windows, and Ubuntu

Installation:
============
The code uses python and matplotlib. If you install anaconda (https://www.continuum.io/downloads), all the necessary components for running the main code should be there.

Mayavi (http://docs.enthought.com/mayavi/mayavi/) was used for visualizing the skeletons. If you used anaconda to install python, you can apply the following command to install Mayavi:

conda install mayavi


Usage Instruction
=================
For common usage check runDemo.sh

Details:
===========
For a web demo and further detail, please visit: 
https://www.cs.rochester.edu/hci/currentprojects.php?proj=automanner

Algorithm and Implementational details:
======================================
M. Iftekhar Tanveer, Ji Liu, M. Ehsan Hoque, "Unsupervised Extraction of Human-Interpretable Nonverbal Behavioral Cues in a Public Speaking Scenario", 23rd ACM Multimedia (ACMMM) Conference, Brisbane, Australia, 2015.

M. Iftekhar Tanveer, Ru Zhao, Kezhen Chen, Zoe Tiet, M. Ehsan Hoque, "AutoManner: An Automated Interface for Making Public Speakers Aware of Their Mannerisms", ACM Intelligent User Interfaces (IUI), Sonoma, CA, USA, 2016.

Requests:
=========
Please cite the papers mentioned above if this code comes to any of your use.

Please share your code if you make any modification to it.