This page shares two datasets of pedestrian positive patches extracted from video sequences captured on a traffic junction at two different days.
This sequences are known as
GSDataset-PANKit, but the patches where extracted from sequences others than GS06 and GS54 datasets.
It is shared the RGB patch of each pedestrian which was enlarged by 6 pixels on with and 6 pixels on height values from its original size.
The Movement Feature Space (MFS) descriptor corresponding to each patch is also include on the distribution.
The MFS information consist on the cummulated level lines (st_gray_mfs_patch folder) and their orientations (ot_gray_mfs_patch folder).
More details on the MFS extraction and use on classification can be found at [1].
The datasets are the following:
- GS_11_09.zip (67 Mbytes, main encripted zip file).
This dataset contains 1,373 colored pedestrian patches.
- GS_13_09.zip (77 Mbytes, main encripted zip file).
This dataset contains 1,990 colored pedestrian patches.
These datasets can be referred as "PedestrianPatchesDataset-PANKit".
Any commercial use of the datasets is strictly forbidden.
Please give appropriate acknowledgements when using these databases and quote
the following paper :
[DYNA2017]
Pablo Negri, Damian Garayalde
"Pedestrian Tracking Using Probability Fields and a Movement Feature Space"
, DYNA Colombia, Vol. 84 (200), pages 217-227, January 2017
[
pdf].
The dataset GS_11_09 and GS_13_09 where recorded under license Nro. DI-2012-1554-DGINC of the 24/08/2012 of the City of Buenos Aires.
This work was supported by PICT-2283 of ANPCyT, CONICET (Argentina), and the ACyT A14T13 of UADE.
References:
[1] P. Negri, N. Goussies, P. Lotito, "Detecting pedestrians on a Movement Feature Space", vol. 47, issue 1, pages 56-71, January 2014.
