iLIDS stands for International Logistic Identification, and is an image dataset for re-identification of occluded pedestrians. The dataset contains 476 images of 119 pedestrians, some of which are occluded by other people or luggage. These images are taken by 4 non-overlapping cameras. The test set contains 238 images, and the validation set contains 238 images.

This dataset can be used to train the learning model FPR (Foreground-aware Pyramid Reconstruction), thereby solving the problem of re-identification of occluded pedestrians.