g3doc/setup/motchallenge_step.md

Run DeepLab2 on MOTChallenge-STEP dataset

MOTChallenge-STEP dataset

MOTChallenge-STEP extends the existing MOTChallenge dataset with spatially and temporally dense annotations.

Label Map

MOTChallenge-STEP dataset followings the same annotation and label policy as KITTI-STEP dataset. Among the MOTChallenge dataset, 4 outdoor sequences are annotated for MOTChallenge-STEP. In particular, these sequences are splitted into 2 for training and 2 for testing. This dataset contains only 7 semantic classes, as not all of Cityscapes' 19 semantic classes are present.

Label Name	Label ID
sidewalk	0
building	1
vegetation	2
sky	3
person†	4
rider	5
bicycle	6
void	255

†: Single instance annotations are available.

Prepare MOTChallenge-STEP for Training and Evaluation

Download data

1. Download MOTChallenge images from https://motchallenge.net/
2. Download groundtruth MOTChallenge-STEP panoptic maps from http://storage.googleapis.com/gresearch/tf-deeplab/data/motchallenge-step.tar.gz

The groundtruth panoptic map is encoded in the same way as described in KITTI-STEP dataset.

Create tfrecord files

You should follow the same folder structure and run the command line script in KITTI-STEP dataset to prepare MOTChallenge-STEP dataset.

Citing MOTChallenge-STEP

If you find this dataset helpful in your research, please use the following BibTeX entry.

@article{step_2021,
  author={Mark Weber and Jun Xie and Maxwell Collins and Yukun Zhu and Paul Voigtlaender and Hartwig Adam and Bradley Green and Andreas Geiger and Bastian Leibe and Daniel Cremers and Aljosa Osep and Laura Leal-Taixe and Liang-Chieh Chen},
  title={{STEP}: Segmenting and Tracking Every Pixel},
  journal={arXiv:2102.11859},
  year={2021}
}