In this page, you’ll find our dataset for several video-surveillance purposes. The dataset is built with purposes of automatically recognition of human activities in video recordings and for social signaling analysis. However, it is also suitable to evaluate and compare other methods related to the video surveillance, such as tracking and detection algorithms.
The dataset contains a set of 12 multi-camera (two cameras) sequences acquired from a real-world scenario at 15 fps, i.e., a vending machines room. We did not hire actors t0 build this dataset. It is a real-life recording. For this reason, we will ask you to sign a form in order to obtain the dataset.
We also provide the full calibration parameters of the two cameras. In order to read the camera data (XML format) and use them, a C++ code is available here. NOTE: The ground plane is assumed to have Z=0.
In order to understand the geometry of the camera-monitored room, the camera views and ground plane positions are reported in the following:
 |
 |
The pyramid corresponds to the cameras (positions + orientations). The yellow plane is the floor, the other planes are the walls of the room.
Some Examples of the sequences:
| Seq 1 Cam 1 | Seq 1 Cam 2 | Seq 6 Cam 1 | Seq 6 Cam 2 |
 |
 |
 |
 |
| Seq 8 Cam 1 | Seq 8 Cam 2 | Seq 9 Cam 1 | Seq 9 Cam 2 |
 |
 |
 |
 |
Download the dataset:
To obtain this dataset, I will ask you some information about your affiliation and your purposes. This is in order to guarantee the privacy and the rights of the recorded people. After that, I will send you the credentials to download it. Note that the dataset is available only for research purposes.
- Fill out this form: PDF
- Send it to: loris dot bazzani at univr dot it (Note: you should send the email from an email address that is linked to your research institution/university)
- Wait for the credentials
- Download the dataset HERE
- Download the calibration parameters HERE (if you need them)
- REMEMBER TO CITE OUR PAPER:
@article{Bazzani:ExpSys12,
author = {Bazzani, L. and Tosato, D. and Cristani, M. and Farenzena, M. and Pagetti, G. and Menegaz, G. and Murino, V.},
journal = {Expert Systems},
note = {in print},
title = {Social Interactions by Visual Focus of Attention in a Three-Dimensional Environment},
year = {2012},
}