ut ut

Laboratory for Image & Video Engineering

Welcome to the IDEAL-LIVE Public-Domain Distorted Face Database

Distorted Face Database (DFD)

Introduction

The IDEAL-LIVE Distorted Face Database (DFD) was created from images available freely on the internet. A total of 215 images were crawled, each with one or more frontal faces. These images were manually ensured to be of high quality with no visible distortions. The images were resized so that the average size of the faces within each image is 80x64 (care was taken to ensure that all the faces within an image are approximately of same size). The faces in these images were manually annotated using 80x64 bounding boxes. This set of 215 images was divided into 150 training images and 65 test images. A total of 1231 faces were present in the training set of images and 393 were present in the test set.

Next, the images were modified in various ways to create distortions. The following distortion types were introduced at multiple severity levels on the training and test datasets:

  • Additive white gaussian noise (AWGN)
  • Gaussian blur (GBlur)
  • JPEG compression (JPEG)

Download

We are making the Distorted Face Database available to the research community free of charge. If you use this database in your research, we kindly ask that you reference our paper listed below:

S. Gunasekar, J. Ghosh, A. C. Bovik, "Face Detection on Distorted Images Augmented by Perceptual Quality-Aware Features," IEEE Transactions on Information Forensics and Security, vol.9, no.12, pp. 2119-2131, December 2014.

Please use THIS LINK to download the database. Follow the instructions in the accompanying README to traverse through the database. Pre-prints of the papers are also available upon request, please contact Suriya Gunasekar ( suriya@utexas.edu ).

Investigators

The investigators on this research were:

Copyright Notice

-----------COPYRIGHT NOTICE STARTS WITH THIS LINE------------
Copyright (c) 2014 The University of Texas at Austin
All rights reserved.

Permission is hereby granted, without written agreement and without license or royalty fees, to use, copy, modify, and distribute this database (the images, the results and the source files) and its documentation for any purpose, provided that the copyright notice in its entirety appear in all copies of this database, and the original source of this database, Laboratory for Image and Video Engineering (LIVE, http://live.ece.utexas.edu ) and Center for Perceptual Systems (CPS, http://www.cps.utexas.edu ) at the University of Texas at Austin (UT Austin, http://www.utexas.edu ), is acknowledged in any publication that reports research using this database.

The following paper is to be cited in the bibliography whenever the database is used as:

S. Gunasekar, J. Ghosh, A. C. Bovik, "Face Detection on Distorted Images Augmented by Perceptual Quality-Aware Features," IEEE Transactions on Information Forensics and Security, vol.9, no.12, pp. 2119-2131, December 2014.

IN NO EVENT SHALL THE UNIVERSITY OF TEXAS AT AUSTIN BE LIABLE TO ANY PARTY FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OF THIS DATABASE AND ITS DOCUMENTATION, EVEN IF THE UNIVERSITY OF TEXAS AT AUSTIN HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

THE UNIVERSITY OF TEXAS AT AUSTIN SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE DATABASE PROVIDED HEREUNDER IS ON AN "AS IS" BASIS, AND THE UNIVERSITY OF TEXAS AT AUSTIN HAS NO OBLIGATION TO PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.

-----------COPYRIGHT NOTICE ENDS WITH THIS LINE------------

Back to Quality Assessment Research page