QMUL Junction Dataset

Back

Download

Download Anomaly Detection Ground-Truth

Details

A busy traffic dataset for research on activity analysis and behaviour understanding.

Video length: 1 hour (90000 frames)
Frame size: 360x288
Frame rate: 25 Hz
Compression codec: ffdshow mpeg-4

The dataset is intended for research purposes only and as such cannot be used commercially.

References

  1. Stream-based Joint Exploration-Exploitation Active Learning
    C. C. Loy, T. M. Hospedales, T. Xiang, and S. Gong
    in Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pp. 1560-1567, 2012 (CVPR, Oral)
     PDF Project Page | Codes
  2. Detecting and Discriminating Behavioural Anomalies
    C. C. Loy, T. Xiang, and S. Gong
    Pattern Recognition, vol. 44, no. 1, pp. 117-132, 2011 (PR)
    PDF
  3. Stream-based Active Unusual Event Detection
    C. C. Loy, T. Xiang, and S. Gong
    in Proceedings of Asian Conference on Computer Vision, pp. 161-175, 2010 (ACCV, Oral)
    PDF Ground-Truth
  4. Modelling Multi-object Activity by Gaussian Processes
    C. C. Loy, T. Xiang, and S. Gong
    British Machine Vision Conference, 2009 (BMVC)
    PDF
  5. From Local Temporal Correlation to Global Anomaly Detection
    C. C. Loy, T. Xiang, and S. Gong
    in Proceedings of European Conference on Computer Vision, International Workshop on Machine Learning for Vision-based Motion Analysis, 2008 (MLVMA @ ECCV)

Acknowledgements

This dataset can also be found here prepared by David Russell and Jian Li with early references.

Other Traffic Datasets