Metadata-Version: 2.4
Name: eval-ab-3d-mot
Version: 2.0.0
Summary: Evaluation for the Pure AB-3D-MOT.
Project-URL: Repository, https://github.com/kovalp/eval-ab-3d-mot
Project-URL: ChangeLog, https://github.com/kovalp/eval-ab-3d-mot/blob/main/CHANGELOG.md
Author-email: Xinshuo Weng <xinshuo.weng@gmail.com>
Maintainer-email: Peter Koval <koval.peter@gmail.com>
License-File: LICENSE
Keywords: ClearMOT,baseline for nuScenes,tracking by detection
Requires-Python: >=3.8
Requires-Dist: binary-classification-ratios>=0.3.0
Requires-Dist: munkres>=1.1.4
Requires-Dist: pure-ab-3d-mot>=2.0.0
Requires-Dist: pytest-mock>=3.14.1
Requires-Dist: rich-argparse>=1.7.1
Requires-Dist: rich>=14.2.0
Requires-Dist: scipy-stubs>=1.14.1.3; python_version > '3.9'
Requires-Dist: scipy>=1.10.1
Description-Content-Type: text/markdown

# Evaluation of a base of 3D multiple-object tracking (AB3DMOT) 

Evaluation part of the AB3DMOT by Xinshuo Weng (https://github.com/xinshuoweng/AB3DMOT)
The purpose of the package is to enable calculation of the detection+tracking quality
metrics for 3D tracking with KITTI data set.

Apart from the refactored evaluation part of the AB3DMOT, a binary classifier 
of the association outcomes is included. The corresponding CLI needs to be implemented still.

## Installation

Should be as easy as `pip install eval-ab-3d-mot`, but if you downloaded the repo,
then `uv sync` standing in the root folder.

## Download the detections & annotations

Should be as easy as

```
git clone https://github.com/kovalp/eval-ab-3d-mot.git
```

The detections (R-CNN) and annotations (training subset of KITTI)
are now in the folder `eval-ab-3d-mot/assets`.

## Command-line scripts

The command-line scripts are equipped with `--help` option which should be 
sufficient to learn their usage.

### Batch run the pure AB-3D-MOT tracker 

```
batch-run-ab-3d-mot assets/detections/kitti/point-r-cnn-training/car/*.txt
```

### Batch evaluation of the pure AB-3D-MOT tracker 

```
batch-eval-ab-3d-mot assets/annotations/kitti/training/*.txt
```
