Skip to content

Latest commit

 

History

History
53 lines (33 loc) · 2.36 KB

README.md

File metadata and controls

53 lines (33 loc) · 2.36 KB
Raw

Data Preparation(Argoverse Tracking v1.0)

Weak Supervision

Use your preferred 2D segmentation method(we made use of inplace-abn) to generate the results for the corresponding sequence and store it following the same file hierarchy. Pre-generated segmentation using inplace_abn for the entire Argoverse Tracking dataset is provided here.

For obtaininig static layouts for entire Argoverse Tracking v1.0 dataset run:

python3 generate_weak_suprevision.py --base_path <path to kitti raw root directory> --seg_class road

GroundTruth

Groundtruth static and dynamic layouts for Argoverse Tracking v1.0 dataset in bird's eye view and can be downloaded from here. The following script generates static and dynamic groundtruth layouts for Argoverse Tracking dataset using the argoverse-api.


# Generating Dynamic Layouts for Argoverse Tracking v1.0 dataset (GroundTruth)
./generate_groundtruth.py --base_path <path to Argoverse Tracking root directory> --seg_class vehicle --range 40 --occ_map_size 256 

# Generating Static Layouts for Argoverse Tracking v1.0 dataset (GroundTruth)
./generate_groundtruth.py --base_path <path to Argoverse Tracking root directory> --seg_class road --range 40 --occ_map_size 256

Usage

usage: generate_argo_groundtruth.py [-h] [--base_path BASE_PATH]
                                    [--out_dir OUT_DIR] [--range RANGE]
                                    [--occ_map_size OCC_MAP_SIZE]
                                    [--seg_class {road,sidewalk,vehicle}]

MonoLayout DataPreparation options

optional arguments:
  -h, --help            show this help message and exit
  --base_path BASE_PATH
                        Path to the root data directory
  --out_dir OUT_DIR     Output directory to save layouts
  --range RANGE         Size of the rectangular grid in metric space
  --occ_map_size OCC_MAP_SIZE
                        Occupancy map size
  --seg_class {road,sidewalk,vehicle}
                        Data Preparation for Road/Sidewalk