iitmcvg / hda_code

Evaluation software to produce perfectly comparable CMC curves and Precision/Recall values on the HDA dataset given a Re-Identification algorithm.

Code of the HDA+ dataset

This is evaluation software to produce perfectly comparable CMC curves and Precision/Recall values with your re-identification algorithm, or produce perfectly comparable MD/FPPI curves for your pedestrian detection algorithm on the HDA+ dataset described here: http://vislab.isr.ist.utl.pt/hda-dataset/

To download this dataset, contact (alex at isr ist utl pt)

Getting Started

• To visualize the dataset with overlaying manual annotations do the following steps:

1. Open HDA_Dataset\hda_code\detection_reid_pipeline\ Computer_Specific_Dataset_Directory.m,
2. Set the hdaRootDirectory variable to the path where you put the HDA_Dataset,
3. Add to the Matlab path the hda_code folder (fill in the addpath(genpath( ... )) ),,
4. and then run MATlab script HDA_Dataset\hda_code\visualization\ RunToVisualize.m

• To see a running example of a sample re-identification algorithm plus the evaluation code, do the following:

1. Open HDA_Dataset\hda_code\detection_reid_pipeline\ Computer_Specific_Dataset_Directory.m,
2. Set the hdaRootDirectory variable to the path where you put the HDA_Dataset,
3. Add to the Matlab path the hda_code folder (fill in the addpath(genpath( ... )) ),,
4. and then run the MATlab script HDA_Dataset\hda_code\detection_reid_pipeline\ fullPipeLineScript.m (more information in hda_code/GettingStarted.txt)

If you use this dataset, please cite:

"The HDA+ data set for research on fully automated re-identification systems." D. Figueira, M. Taiana, A. Nambiar, J. Nascimento and A. Bernardino, VS-RE-ID Workshop at ECCV 2014.

~ Prerequisites: This dataset requires Matlab (TM).