This code has only been tested on the following combination of major pre-requisites. Please check beforehand.
- Ubuntu 22.04
- Python 3.8
To install required dependencies on the system python, please run the following command at the root of this code:
pip install torch==1.12.1+cu113 torchvision==0.13.1+cu113 torchaudio==0.12.1 --extra-index-url https://download.pytorch.org/whl/cu113
pip3 install -r requirements.txt
Download the SemanticKITTI dataset with labels from here.
Add the semantic-kitti.yaml file to the folder.
Folder structure:
SemanticKitti/
└── semantic-kitti.yaml
└── sequences/
└── 00/
└── calib.txt
└── poses.txt
└── times.txt
└── labels
├── 000000.label
├── 000000.center.npy
...
└── velodyne
├── 000000.bin
...
We provided 4D-stop sample result to this repository, so there is no need to run 4DSTOP network for our tracker. just run the two files:
- bounding box tracker :
tracking/box_tracker.py
- from bounding box to labels tracker:
tracking/box_tracker_to_labels.py
if you dont want to run this sample you can run from your own path by :
- bounding box tracker :
tracking/box_tracker.py --dataset [path to the SemanticKitti point cloud] --data_cfg [path to SemanticKitti config file] --predictions [path to prediction] --sequences [sequence number] --split [valid or not]
- from bounding box to labels tracker:
tracking/box_tracker_to_labels.py --dataset [path to the SemanticKitti point cloud] --data_cfg [path to SemanticKitti config file] --predictions [path to prediction] --sequences [sequence number] --split [valid or not]
utils/evaluate_4dpanoptic.py --dataset [path to the SemanticKitti point cloud] --predictions [path to NextStop prediction folder]] --data_cfg [path to SemanticKitti config file] --split [valid or not] --min_inst_points [50 or 1]
The code is based on 4D-StOP, semantic-kitti-api and AB3DMOT