Welcome to Violent-Flows - Crowd Violence \ Non-violence Database and benchmark: A database of real-world, video footage of crowd violence, along with standard benchmark protocols designed to test both violent/non-violent classification and violence outbreak detections. The data set contains 246 videos. All the videos were downloaded from YouTube. The shortest clip duration is 1.04 seconds, the longest clip is 6.52 seconds, and the average length of a video clip is 3.60 seconds.


New! May, 25th, 2014: Code for computing the Violent Flows (ViF) descriptors is now available.

To run, call VIF_create_feature_vec.m with path and file_name of an AVI file. The output is a vector of VIF features.

Download the database

Please provide the details here. The reason we ask for them is that we would like to keep in touch in case we find any need for a critical update or in case we would organize a dedicated workshop, etc. When done, you will be able to see the FTP password. Thanks for your cooperation.

The file Movies.zip is available from www.cslab.openu.ac.il/download. You will need the username and password provided by the form above.


The Violent-Flows Results page.


If you use this database, or refer to its results, please cite the following paper:

T. Hassner, Y. Itcher, and O. Kliper-Gross, Violent Flows: Real-Time Detection of Violent Crowd Behavior, 3rd IEEE International Workshop on Socially Intelligent Surveillance and Monitoring (SISM) at the IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), Rhode Island, June 2012 .


Tal Hassner
Yossi Itcher - yossi.itcher@gmail.com
Orit Kliper-Gross - orit.kliper@weizmann.ac.il