tracktor

Tracking without Bells and Whistles

Introduction

[ALGORITHM]

@inproceedings{bergmann2019tracking,
  title={Tracking without bells and whistles},
  author={Bergmann, Philipp and Meinhardt, Tim and Leal-Taixe, Laura},
  booktitle={Proceedings of the IEEE international conference on computer vision},
  pages={941--951},
  year={2019}
}

We implement Tracktor with independent detector and ReID models. To train a model by yourself, you need to train a detector following here and also train a ReID model following here. The configs in this folder are basiclly for inference.

Results and models on MOT15

Detector	ReID	Train Set	Test Set	Public	Inf time (fps)	MOTA	IDF1	FP	FN	IDSw.	Config	Download
R50-FasterRCNN-FPN	R50	half-train	half-val	Y	-	61.8	64.9	1235	6877	116	config	detector | detector_log | reid | reid_log
R50-FasterRCNN-FPN	R50	half-train	half-val	N	-	66.8	68.4	3049	3922	179	config	detector | detector_log | reid | reid_log

Results and models on MOT16

Detector	ReID	Train Set	Test Set	Public	Inf time (fps)	MOTA	IDF1	FP	FN	IDSw.	Config	Download
R50-FasterRCNN-FPN	R50	half-train	half-val	Y	-	54.1	61.5	425	23894	182	config	detector | detector_log | reid | reid_log
R50-FasterRCNN-FPN	R50	half-train	half-val	N	-	63.4	66.2	4175	14911	444	config	detector | detector_log | reid | reid_log

Results and models on MOT17

The implementations of Tracktor follow the offical practices. In the table below, the result marked with * (the last line) is the offical one. Our implementation outperform it by 4.9 points on MOTA and 3.3 points on IDF1.

Detector	ReID	Train Set	Test Set	Public	Inf time (fps)	MOTA	IDF1	FP	FN	IDSw.	Config	Download
R50-FasterRCNN-FPN	R50	half-train	half-val	Y	3.2	57.3	63.4	1254	67091	614	config	detector reid
R50-FasterRCNN-FPN	R50	half-train	half-val	N	3.1	64.1	66.9	11088	45762	1233	config	detector reid
R50-FasterRCNN-FPN	R50	train	test	Y	3.2	61.2	58.4	8609	207627	2634	config	detector reid
R50-FasterRCNN-FPN*	R50	train	test	Y	-	56.3	55.1	8866	235449	1987	-	-

Results and models on MOT20

The implementations of Tracktor follow the offical practices. In the table below, the result marked with * (the last line) is the offical one. Our implementation outperform it by 5.3 points on MOTA and 2.1 points on IDF1.

Detector	ReID	Train Set	Test Set	Public	Inf time (fps)	MOTA	IDF1	FP	FN	IDSw.	Config	Download
R50-FasterRCNN-FPN	R50	half-train	half-val	Y	-	70.6	65.4	3652	175955	1441	config	detector | detector_log | reid | reid_log
R50-FasterRCNN-FPN	R50	half-train	half-val	N	-	70.9	64.1	5539	171653	1619	config	detector | detector_log | reid | reid_log
R50-FasterRCNN-FPN	R50	train	test	Y	-	57.9	54.8	16203	199485	2299	config	detector | detector_log | reid | reid_log
R50-FasterRCNN-FPN*	R50	train	test	Y	-	52.6	52.7	6930	236680	1648	-	-

Note: When running demo_mot.py, we suggest you use the config containing private, since private means the MOT method doesn't need external detections.