Anomalous Behavior Data Set
Andrei Zaharescu and Richard P. Wildes
Contact email andreiz at cse dot yorku dot ca
Department of Computer Science and Engineering
and Centre for Vision Research
York University, Toronto, ON, Canada
This website provides a data set for anomalous behaviour detection in video. The data set contains 8 image sequences that depict a wide range of challenging scenarios, including: illumination effects, scene clutter, variable target appearance, rapid motion and camera jitter. All sequences are available with manually constructed ground truth that identifies anomalous behaviour relative to a training portion of the video. Also provided is software for groundtruth construction and subsequent evaluation.
For details of our technical approach to anomalous behaviour detection, see our project page.
A number of videos with groundtruth are provided below:
Groundruth and Evaluation Software
In addition, we provide both a groundtruthing software and an evaluation software, in order to both groundtruth future datasets and evaluate other methods against the current data. Please dowload the software (both Windows binaries and source files) from here (they are written in C and they use OpenCV). Consult the provided README file for more information.
If you end up using the software, please let me know about it reference our work appropriately.
- A. Zaharescu and R.P. Wildes, Anomalous Behaviour Detection using Spatiotemporal Oriented Energies, Subset Inclusion Histogram Comparison and Event-Dirven Processing, ECCV, 2010.
- K.G. Derpanis and R.P. Wildes, Dynamic Texture Recognition based on Distributions of Spacetime Oriented Structure, CVPR, 2010.
- K. Derpanis and J. Gryn, Three-dimensional nth derivative of Gaussian
separable steerable filters. ICIP 3, pp.553-556, 2005