<

Example frames. (1) Student Orientation, (2) Group Studying, (3) Career Fair, (4) Forum on Gun Control and Gun Violence, (5) Scholarship Competition, (6) Cleaning.

Download
Including image sequences in jpeg format. A campus event schedule can be found in the table below. To facilitate downloading this huge video dataset, you may want to download and install the Baidu cloud bulter.

Details

The Educational Resource Centre (ERCe) dataset was collected from a publicly accessible webcam deployed on a university campus across about 2 months for semantic event based video synopsis research. This dataset can be exploited for other computer vision tasks, such as video classification, annotation or clustering.

Image sequences: organised by multiple folders named with the record time (GMT time zone). In each folder 3200 frames are contained.

Campus event schedule: mainly extracted from a university event calendar, including Student Orientation (Stud. Orient.), Career Fair, Forum on Gun Control and Gun Violence (Gun Forum), Group Studying, Scholarship Competition (Schlr. Comp.), Accommodative Service (Accom. Service), Cleaning. The detailed timetable of these events is given in the following table (Note that there is a gap of 5 hours between the university's time (UTC-05:00) and GMT time):

Event Date or time
Student Orientation 08-09 January
Group Studying 27 January
Career Fair 07 February
Forum on Gun Control and Gun Violence 08 February
Scholarship Competition 02 March
Cleaning around 05:00-09:00am
Accommodative Service 10:00-10:45am, 14 January
09:30-10:15am, 15 January
10:00-10:45am, 21 January
09:30-10:15am, 22 January
14:00-14:45pm, 23 January
14:30-15:15pm, 24 January
08:00-08:45am, 25 January
14:00-14:45pm, 04 February
09:30-10:15am, 05 February
12:00-12:45pm, 06 February
09:30-10:15am, 12 February
14:00-14:45pm, 15 February


Stream length: a total of 438 folders, each containing 3200 frames
Frame size: 480x640
Frame rate: ~5 Hz

The dataset is intended for research purposes only and as such cannot be used commercially. Please cite the following publication when this dataset is used in any academic and research reports.

References

  1. Video Synopsis by Heterogeneous Multi-Source Correlation
    X. Zhu, C. C. Loy, S. Gong
    in Proceedings of IEEE International Conference on Computer Vision, 2013 (ICCV)
    PDF
  2. Learning from Multiple Sources for Video Summarisation
    X. Zhu, C. C. Loy, S. Gong
    International Journal of Computer Vision, Vol. 117, No. 3, pp. 247-268, May 2016 (IJCV)
    PDF