Faculty of Informatics Vienna University of Technology Institute of Computer Aided Automation PRIP Home PRIP Home
Personal tools
You are here: Home People Julian Stöttinger FeEval

FeEval - A dataset for evaluation of spatio-temporal local features

FeEval consists of 1710 videos based on 30 originals. The videos are from a great variety of different sources, including HDTV shows, 1080p HD movies and surveillance cameras. The videos are iteratively varied by increasing blur, noise, increasing or decreasing light, median filter, codec quality, scale and rotation leading to a total of 1710 video clips. Homography matrices are provided for geometric transformations. The surveillance videos are taken from 4 different angles in a calibrated environment. Similar to prior work on 2D images, this leads to a repeatability and matching measurement in videos for spatio-temporal features estimating the overlap of features under increasing changes in the data.

 

FeEval TV show for evaluation of spatio-temporal features

FeEval video 1 iverview
visible transformations

HDTV shows under well defined transformations to measure robustness of spatio-temporal features.

FeEval Video 14

Surveillance videos in a calibrated environment.

1080p HD movies.

Please reference the database if you find it useful:

FeEval - A dataset for evaluation of spatio-temporal local features

Julian Stöttinger, Sebastian Zambanini, Rehanulla Khan, Allan Hanbury, Proceedings of the 20th International Conference on Pattern Recognition (ICPR), Istanbul, Turkey, August 23-26, 2010, to appear.

FeEval download

Homography matrices and source code is available here.

Some of the material is copyrighted. It is provided for scientific use only.