JRDB Leaderboard

2D Tracking Submissions

Name

MOTA↑

OSPA⁽²⁾_IoU ↓

IDF1↑

HOTA↑

Tracktor++_yw

32.414%

0.92

28.018%

25.24%

Anonymous Submission

Details

Name	Tracktor++_yw
Submission Date	2021-12-16 12:29:20+00:00
Abstract	The problem of tracking multiple objects in a video se- quence poses several challenging tasks. For tracking-by- detection, these include object re-identification, motion pre- diction and dealing with occlusions. We present a tracker (without bells and whistles) that accomplishes tracking without specifically targeting any of these tasks, in partic- ular, we perform no training or optimization on tracking data. To this end, we exploit the bounding box regression of an object detector to predict the position of an object in the next frame, thereby converting a detector into a Tracktor. We demonstrate the potential of Tracktor and provide a new state-of-the-art on three multi-object tracking benchmarks by extending it with a straightforward re-identification and camera motion compensation. We then perform an analysis on the performance and failure cases of several state-of-the-art tracking methods in comparison to our Tracktor. Surprisingly, none of the dedicated tracking methods are considerably better in deal- ing with complex tracking scenarios, namely, small and occluded objects or missing detections. However, our ap- proach tackles most of the easy tracking scenarios. There- fore, we motivate our approach as a new tracking paradigm and point out promising future research directions. Over- all, Tracktor yields superior tracking performance than any current tracking method and our analysis exposes remain- ing and unsolved tracking challenges to inspire future re- search directions.
Publication title	N/A	Publication authors:	N/A
Publication venue and year	N/A	Publication URL:	N/A
Code Language	N/A	Hardware	1 GPU (Titan X)
Code Website	N/A	Code URL	N/A

Visualisation

hewlett-class-2019-01-23_1

Best 3 By MOTA ↓

Worst 3 By MOTA ↓

Per-sequence Results

Sequence	OSPA	OSPA_CARD	OSPA_LOC
cubberly-auditorium-2019-04-22_1	0.88123	0.58242	0.29881
discovery-walk-2019-02-28_0	0.96016	0.73737	0.22279
discovery-walk-2019-02-28_1	0.95189	0.69841	0.25347
food-trucks-2019-02-12_0	0.96366	0.79239	0.17127
gates-ai-lab-2019-04-17_0	0.98607	0.9257	0.060376
gates-basement-elevators-2019-01-17_0	0.85727	0.66387	0.1934
gates-foyer-2019-01-17_0	0.92604	0.82051	0.10553
gates-to-clark-2019-02-28_0	0.90152	0.69048	0.21105
hewlett-class-2019-01-23_0	0.86818	0.61236	0.25582
hewlett-class-2019-01-23_1	0.69728	0.46296	0.23432
huang-2-2019-01-25_1	0.92189	0.67797	0.24393
huang-intersection-2019-01-22_0	0.9618	0.66538	0.29641
indoor-coupa-cafe-2019-02-06_0	0.98064	0.83908	0.14156
lomita-serra-intersection-2019-01-30_0	0.9653	0.73723	0.22807
meyer-green-2019-03-16_1	0.96625	0.70909	0.25716
nvidia-aud-2019-01-25_0	0.96177	0.80923	0.15253
nvidia-aud-2019-04-18_1	0.86614	0.51786	0.34829
nvidia-aud-2019-04-18_2	0.8413	0.44262	0.39868
quarry-road-2019-02-28_0	0.96495	0.81102	0.15392
serra-street-2019-01-30_0	0.98262	0.81626	0.16636
stlc-111-2019-04-19_1	0.80275	0.57895	0.22381
stlc-111-2019-04-19_2	0.86701	0.64151	0.2255
tressider-2019-03-16_2	0.95353	0.58955	0.36398
tressider-2019-04-26_0	0.96641	0.7478	0.21861
tressider-2019-04-26_1	0.97056	0.74594	0.22462
tressider-2019-04-26_3	0.95826	0.73913	0.21913
COMBINED	0.92017	0.69443	0.22575

Sequence	HOTA	DetA	AssA	DetRe	DetPr	AssRe	AssPr	LocA
cubberly-auditorium-2019-04-22_1	33.82	37.092	31.131	39.641	72.464	33.677	72.332	79.841
discovery-walk-2019-02-28_0	28.243	30.421	27.674	32.122	73.796	29.258	79.135	80.335
discovery-walk-2019-02-28_1	29.77	33.053	26.966	35.001	73.83	28.46	74.715	80.507
food-trucks-2019-02-12_0	30.055	38.234	24.04	39.725	79.999	25.013	78.572	82.659
gates-ai-lab-2019-04-17_0	27.123	41.273	17.958	46.155	72.216	18.436	80.364	83.683
gates-basement-elevators-2019-01-17_0	40.095	45.756	35.434	48.13	78.646	38.459	77.684	82.139
gates-foyer-2019-01-17_0	38.376	53.735	27.49	62.159	71.711	28.681	81.589	83.58
gates-to-clark-2019-02-28_0	37.023	48	28.626	54.251	73.07	32.359	76.269	83.979
hewlett-class-2019-01-23_0	36.762	49.181	27.627	52.322	79.766	32.41	65.483	83.789
hewlett-class-2019-01-23_1	58.274	69.537	48.884	74.114	82.247	52.172	75.747	84.907
huang-2-2019-01-25_1	33.354	34.829	32.211	38.789	69.151	34.558	73.112	82.283
huang-intersection-2019-01-22_0	21.681	22.011	22.305	23.553	65.31	23.64	69.519	77.741
indoor-coupa-cafe-2019-02-06_0	19.363	28.236	13.46	29.612	74.949	13.994	75.826	81.067
lomita-serra-intersection-2019-01-30_0	29.713	23.644	37.753	25.552	65.925	39.546	71.567	79.278
meyer-green-2019-03-16_1	23.303	21.383	25.7	22.76	66.523	27.241	71.732	78.218
nvidia-aud-2019-01-25_0	26.84	32.924	22.674	35.319	70.305	24.471	74.277	78.668
nvidia-aud-2019-04-18_1	48.521	43.251	55.484	46.767	77.21	57.912	86.494	84.197
nvidia-aud-2019-04-18_2	40.506	42.179	39.252	46.089	70.589	42.634	68.802	79.394
quarry-road-2019-02-28_0	29.004	25.923	32.573	30.276	55.07	34.494	73.126	76.799
serra-street-2019-01-30_0	21.224	17.772	26.296	18.931	63.784	27.929	73.344	77.948
stlc-111-2019-04-19_1	53.037	56.188	50.216	61.389	77.821	52.956	80.253	84.002
stlc-111-2019-04-19_2	42.983	54.966	33.673	61.63	74.796	35.868	74.785	83.518
tressider-2019-03-16_2	23.348	18.896	29.895	19.598	70.869	32.454	69.129	78.339
tressider-2019-04-26_0	17.686	21.907	14.516	22.942	70.728	15.415	69.818	78.887
tressider-2019-04-26_1	21.207	25.047	18.202	26.021	75.343	19.534	70.246	80.658
tressider-2019-04-26_3	20.166	27.639	15.231	29.677	67.593	16.552	65.799	77.922
COMBINED	25.24	28.624	22.688	30.367	72.342	24.259	72.921	80.369

Sequence	MOTA	MOTP	MODA	CLR_Re	CLR_Pr	MTR	PTR	sMOTA	CLR_TP	CLR_FN	CLR_FP	IDSW	MT	PT	ML
cubberly-auditorium-2019-04-22_1	46.225	23.86	46.799	50.752	92.774	13.158	63.158	7.5823	6715	6516	523	76	5	24	9
discovery-walk-2019-02-28_0	34.57	23.032	35.373	39.45	90.633	7.6923	53.846	4.2058	5254	8064	543	107	2	14	10
discovery-walk-2019-02-28_1	38.814	22.741	39.484	43.446	91.642	10.526	50	5.2481	6294	8193	574	97	4	19	15
food-trucks-2019-02-12_0	47.104	20.4	47.604	48.63	97.933	18.333	46.667	8.3949	31171	32927	658	320	11	28	21
gates-ai-lab-2019-04-17_0	43.097	18.161	43.736	53.824	84.216	29.167	37.5	-0.95212	13899	11924	2605	165	7	9	8
gates-basement-elevators-2019-01-17_0	56.812	21.254	57.49	59.344	96.97	42.5	52.5	10.082	7361	5043	230	84	17	21	2
gates-foyer-2019-01-17_0	62.367	18.964	62.868	74.774	86.264	38.095	52.381	1.7738	7618	2570	1213	51	8	11	2
gates-to-clark-2019-02-28_0	53.273	18.301	54.175	64.21	86.484	53.846	23.077	0.81393	1638	913	256	23	7	3	3
hewlett-class-2019-01-23_0	59.534	18.823	60.599	63.096	96.192	34.783	62.319	8.3143	10433	6102	413	176	24	43	2
hewlett-class-2019-01-23_1	84.252	17.067	84.855	87.483	97.084	72.414	24.138	11.699	4494	643	135	31	21	7	1
huang-2-2019-01-25_1	38.084	20.747	38.537	47.315	84.351	15.789	57.895	0.58572	3137	3493	582	30	3	11	5
huang-intersection-2019-01-22_0	22.523	25.803	23.097	29.58	82.022	2.2989	49.425	0.57564	13870	33020	3040	269	2	43	42
indoor-coupa-cafe-2019-02-06_0	32.392	21.725	32.832	36.171	91.55	3.5714	53.571	4.0788	23727	41870	2190	289	2	30	24
lomita-serra-intersection-2019-01-30_0	22.672	23.318	23.182	30.97	79.905	5.5556	50	-1.0768	6378	14216	1604	105	2	18	16
meyer-green-2019-03-16_1	22.017	24.893	22.637	28.425	83.082	3.125	40.625	0.66797	6654	16755	1355	145	2	26	36
nvidia-aud-2019-01-25_0	36.182	24.534	36.919	43.578	86.744	14.516	51.613	3.2957	14436	18691	2206	244	9	32	21
nvidia-aud-2019-04-18_1	46.956	17.719	47.345	53.958	89.082	29.63	25.926	2.5593	4308	3676	528	31	8	7	12
nvidia-aud-2019-04-18_2	52.485	24.946	52.956	59.124	90.554	29.412	41.176	8.11	6020	4162	628	48	10	14	10
quarry-road-2019-02-28_0	21.039	27.269	21.699	38.338	69.735	8.3333	37.5	-6.8441	2265	3643	983	39	2	9	13
serra-street-2019-01-30_0	15.098	24.989	16.365	23.023	77.569	3.5398	43.363	-2.1714	9904	33114	2864	545	4	49	60
stlc-111-2019-04-19_1	66.149	18.031	66.773	72.829	92.323	45.833	50	6.4511	4546	1696	378	39	11	12	1
stlc-111-2019-04-19_2	62.617	18.627	63.609	73.003	88.599	31.579	52.632	3.2126	2650	980	341	36	6	10	3
tressider-2019-03-16_2	21.842	25.295	22.289	24.971	90.3	7.2727	29.091	3.1875	6656	19999	715	119	4	16	35
tressider-2019-04-26_0	24.365	24.089	25.229	28.833	88.889	2.7132	51.938	2.4777	34520	85205	4315	1034	7	134	117
tressider-2019-04-26_1	29.695	22.119	30.264	32.4	93.813	7.3059	37.443	4.4618	62383	130155	4114	1094	16	82	121
tressider-2019-04-26_3	30.614	25.338	31.54	37.723	85.918	5.9829	52.991	2.449	44003	72644	7212	1081	14	124	96
COMBINED	32.414	22.582	33.107	37.542	89.435	12.315	47.128	3.3503	340334	566214	40205	6278	208	796	685

Sequence	IDF1	IDR	IDP	IDTP	IDFN	IDFP
cubberly-auditorium-2019-04-22_1	44.936	34.759	63.54	4599	8632	2639
discovery-walk-2019-02-28_0	31.452	22.571	51.854	3006	10312	2791
discovery-walk-2019-02-28_1	37.303	27.494	57.994	3983	10504	2885
food-trucks-2019-02-12_0	32.491	24.313	48.962	15584	48514	16245
gates-ai-lab-2019-04-17_0	25.019	20.505	32.083	5295	20528	11209
gates-basement-elevators-2019-01-17_0	50.483	40.688	66.487	5047	7357	2544
gates-foyer-2019-01-17_0	41.821	39.036	45.035	3977	6211	4854
gates-to-clark-2019-02-28_0	36.895	32.144	43.295	820	1731	1074
hewlett-class-2019-01-23_0	44.571	36.904	56.26	6102	10433	4744
hewlett-class-2019-01-23_1	68.667	65.272	72.435	3353	1784	1276
huang-2-2019-01-25_1	42.4	33.092	58.994	2194	4436	1525
huang-intersection-2019-01-22_0	25.781	17.539	48.634	8224	38666	8686
indoor-coupa-cafe-2019-02-06_0	22.81	15.911	40.271	10437	55160	15480
lomita-serra-intersection-2019-01-30_0	34.567	23.983	61.877	4939	15655	3043
meyer-green-2019-03-16_1	27.265	18.296	53.477	4283	19126	3726
nvidia-aud-2019-01-25_0	31.879	23.947	47.669	7933	25194	8709
nvidia-aud-2019-04-18_1	53.448	42.911	70.844	3426	4558	1410
nvidia-aud-2019-04-18_2	49.424	40.847	62.56	4159	6023	2489
quarry-road-2019-02-28_0	40.017	31.009	56.404	1832	4076	1416
serra-street-2019-01-30_0	19.865	12.881	43.398	5541	37477	7227
stlc-111-2019-04-19_1	67.616	60.477	76.665	3775	2467	1149
stlc-111-2019-04-19_2	52.802	48.154	58.442	1748	1882	1243
tressider-2019-03-16_2	24.793	15.824	57.224	4218	22437	3153
tressider-2019-04-26_0	23.804	15.763	48.595	18872	100853	19963
tressider-2019-04-26_1	21.812	14.673	42.485	28251	164287	38246
tressider-2019-04-26_3	22.289	16.037	36.526	18707	97940	32508
COMBINED	28.018	19.889	47.381	180305	726243	200234

Sequence	Dets	IDs
cubberly-auditorium-2019-04-22_1	7238	91
discovery-walk-2019-02-28_0	5797	99
discovery-walk-2019-02-28_1	6868	126
food-trucks-2019-02-12_0	31829	289
gates-ai-lab-2019-04-17_0	16504	323
gates-basement-elevators-2019-01-17_0	7591	119
gates-foyer-2019-01-17_0	8831	117
gates-to-clark-2019-02-28_0	1894	42
hewlett-class-2019-01-23_0	10846	178
hewlett-class-2019-01-23_1	4629	54
huang-2-2019-01-25_1	3719	59
huang-intersection-2019-01-22_0	16910	260
indoor-coupa-cafe-2019-02-06_0	25917	348
lomita-serra-intersection-2019-01-30_0	7982	137
meyer-green-2019-03-16_1	8009	220
nvidia-aud-2019-01-25_0	16642	325
nvidia-aud-2019-04-18_1	4836	56
nvidia-aud-2019-04-18_2	6648	61
quarry-road-2019-02-28_0	3248	127
serra-street-2019-01-30_0	12768	615
stlc-111-2019-04-19_1	4924	57
stlc-111-2019-04-19_2	2991	53
tressider-2019-03-16_2	7371	134
tressider-2019-04-26_0	38835	1023
tressider-2019-04-26_1	66497	862
tressider-2019-04-26_3	51215	897
COMBINED	380539	6672

MMPAT_CVPR21

32.326%

0.886

33.691%

29.931%

Y. He, W. Yu, J. Han, X. Wei, X. Hong and Y. Gong Know Your Surroundings: Panoramic Multi-Object Tracking by Multimodality Collaboration in CVPRW, 2021

Details

Name	MMPAT_CVPR21
Submission Date	2021-12-16 12:21:38+00:00
Abstract	Panoramic image detection and tracking using both image and point clouds. For tracking, we adopt the Cascade rcnn as our object detector. The backbone of the detector is ResNet 101. The detector is pre-trained on the COCO dataset and then fine tuned on the JRDB dataset. During training, we employ data augment methods such as mixup, random crop, multiscale training to augment the dataset. During inference, we employ horizontal flip and softnms techniques to generate more robust resutls. For tracking, we fuse both the 3D image and 3D point clouds for robust tracking.
Publication title	Know Your Surroundings: Panoramic Multi-Object Tracking by Multimodality Collaboration	Publication authors:	Y. He, W. Yu, J. Han, X. Wei, X. Hong and Y. Gong
Publication venue and year	CVPRW, 2021	Publication URL:	https://arxiv.org/abs/2105.14683
Code Language	N/A	Hardware	1 GPU (Titan X)
Code Website	N/A	Code URL	N/A

Visualisation

hewlett-class-2019-01-23_1

Best 3 By MOTA ↓

Worst 3 By MOTA ↓

Per-sequence Results

Sequence	OSPA	OSPA_CARD	OSPA_LOC
cubberly-auditorium-2019-04-22_1	0.86856	0.55814	0.31042
discovery-walk-2019-02-28_0	0.94322	0.73469	0.20853
discovery-walk-2019-02-28_1	0.92046	0.66667	0.25379
food-trucks-2019-02-12_0	0.94075	0.74359	0.19716
gates-ai-lab-2019-04-17_0	0.97581	0.87097	0.10484
gates-basement-elevators-2019-01-17_0	0.81564	0.56989	0.24575
gates-foyer-2019-01-17_0	0.89923	0.74074	0.15849
gates-to-clark-2019-02-28_0	0.87909	0.675	0.20409
hewlett-class-2019-01-23_0	0.849	0.58929	0.25971
hewlett-class-2019-01-23_1	0.68574	0.46296	0.22277
huang-2-2019-01-25_1	0.91579	0.66071	0.25508
huang-intersection-2019-01-22_0	0.91233	0.53476	0.37757
indoor-coupa-cafe-2019-02-06_0	0.97623	0.79783	0.1784
lomita-serra-intersection-2019-01-30_0	0.91424	0.59551	0.31873
meyer-green-2019-03-16_1	0.9208	0.49606	0.42474
nvidia-aud-2019-01-25_0	0.9335	0.72687	0.20663
nvidia-aud-2019-04-18_1	0.7892	0.38636	0.40283
nvidia-aud-2019-04-18_2	0.74307	0.34615	0.39691
quarry-road-2019-02-28_0	0.90459	0.61905	0.28555
serra-street-2019-01-30_0	0.86615	0.38919	0.47696
stlc-111-2019-04-19_1	0.81191	0.58621	0.22571
stlc-111-2019-04-19_2	0.81714	0.525	0.29214
tressider-2019-03-16_2	0.93642	0.64052	0.2959
tressider-2019-04-26_0	0.95711	0.69754	0.25957
tressider-2019-04-26_1	0.94028	0.60682	0.33345
tressider-2019-04-26_3	0.93166	0.6256	0.30606
COMBINED	0.88646	0.60947	0.27699

Sequence	HOTA	DetA	AssA	DetRe	DetPr	AssRe	AssPr	LocA
cubberly-auditorium-2019-04-22_1	35.795	39.383	32.829	42.974	73.576	35.777	75.018	82.327
discovery-walk-2019-02-28_0	29.958	35.539	26.475	41.492	63.204	28.626	71.943	80.311
discovery-walk-2019-02-28_1	37.584	43.241	33.031	47.419	75.44	34.509	81.354	83.789
food-trucks-2019-02-12_0	36.02	41.935	31.313	44.392	79.868	32.27	83.925	84.343
gates-ai-lab-2019-04-17_0	27.414	40.363	18.716	45.463	73.177	19.497	83.506	86.251
gates-basement-elevators-2019-01-17_0	41.92	44.254	39.86	46.177	83.838	42.284	83.422	86.252
gates-foyer-2019-01-17_0	38.898	55.175	27.486	66.075	71.881	29.075	84.986	86.378
gates-to-clark-2019-02-28_0	39.808	47.988	33.041	56.882	71.375	34.125	87.619	87.178
hewlett-class-2019-01-23_0	41.482	44.598	38.824	49.147	76.261	42.541	81.353	85.59
hewlett-class-2019-01-23_1	64.618	69.999	59.778	74.525	84.605	62.256	86.447	86.926
huang-2-2019-01-25_1	34.406	38.041	32.572	44.982	64.012	35.705	76.959	81.676
huang-intersection-2019-01-22_0	29.774	28.547	31.638	32.26	62.647	34.511	70.571	79.541
indoor-coupa-cafe-2019-02-06_0	20.473	29.175	14.554	31.023	74.001	14.876	80.913	82.097
lomita-serra-intersection-2019-01-30_0	35.244	29.04	44.952	33.903	59.36	47.489	79.039	79.611
meyer-green-2019-03-16_1	28.266	24.488	33.813	26.685	66.616	36.157	74.352	80.606
nvidia-aud-2019-01-25_0	32.504	37.231	29.229	40.338	72.06	30.589	79.612	80.403
nvidia-aud-2019-04-18_1	52.766	47.468	59.384	51.737	77.008	61.776	85.43	83.868
nvidia-aud-2019-04-18_2	53.428	47.363	60.475	52.138	72.493	63.329	82.032	81.139
quarry-road-2019-02-28_0	34.449	34.92	34.377	42.734	59.031	37.138	71.054	80.599
serra-street-2019-01-30_0	30.438	24.736	38.882	27.195	63.1	42.706	71.066	78.163
stlc-111-2019-04-19_1	52.332	58.739	46.893	64.757	80.923	49.552	84.24	87.278
stlc-111-2019-04-19_2	47.032	55.432	39.966	62.964	76.93	44.206	77.773	86.606
tressider-2019-03-16_2	30.791	26.537	37.637	30.765	58.077	40.261	72.196	78.938
tressider-2019-04-26_0	19.473	23.096	16.746	24.38	71.197	17.656	74.419	80.329
tressider-2019-04-26_1	27.628	26.66	28.891	27.675	79.065	30.064	80.301	83.52
tressider-2019-04-26_3	23.588	27.91	20.461	30.256	68.264	21.739	75.332	79.86
COMBINED	29.931	31.093	29.305	33.661	71.914	30.966	78.876	82.135

Sequence	MOTA	MOTP	MODA	CLR_Re	CLR_Pr	MTR	PTR	sMOTA	CLR_TP	CLR_FN	CLR_FP	IDSW	MT	PT	ML
cubberly-auditorium-2019-04-22_1	45.061	21.119	45.922	52.165	89.312	18.421	57.895	3.9125	6902	6329	826	114	7	22	9
discovery-walk-2019-02-28_0	28.54	21.233	29.877	47.762	72.755	11.538	57.692	-9.0806	6361	6957	2382	178	3	15	8
discovery-walk-2019-02-28_1	45.524	17.943	46.29	54.573	86.822	13.158	57.895	0.74264	7906	6581	1200	111	5	22	11
food-trucks-2019-02-12_0	49.207	18.071	49.588	52.585	94.608	15	58.333	6.1248	33706	30392	1921	244	9	35	16
gates-ai-lab-2019-04-17_0	39.98	15.2	40.503	51.315	82.597	25	41.667	-3.5352	13251	12572	2792	135	6	10	8
gates-basement-elevators-2019-01-17_0	51.548	15.797	52.064	53.571	97.263	40	50	6.4392	6645	5759	187	64	16	20	4
gates-foyer-2019-01-17_0	58.255	15.368	58.785	75.353	81.975	42.857	47.619	-5.5184	7677	2511	1688	54	9	10	2
gates-to-clark-2019-02-28_0	45.472	13.93	46.452	63.073	79.144	30.769	46.154	-8.8148	1609	942	424	25	4	6	3
hewlett-class-2019-01-23_0	50.172	16.824	50.934	57.69	89.518	27.536	66.667	2.1886	9539	6996	1117	126	19	46	4
hewlett-class-2019-01-23_1	81.877	14.87	82.48	85.283	96.818	68.966	27.586	9.2746	4381	756	144	31	20	8	1
huang-2-2019-01-25_1	36.787	21.345	37.451	53.861	76.647	21.053	47.368	-5.577	3571	3059	1088	44	4	9	6
huang-intersection-2019-01-22_0	25.234	22.875	25.942	38.718	75.188	3.4483	63.218	-4.6279	18155	28735	5991	332	3	55	29
indoor-coupa-cafe-2019-02-06_0	31.608	20.562	32.078	37	88.258	3.5714	50	2.2161	24271	41326	3229	308	2	28	26
lomita-serra-intersection-2019-01-30_0	13.902	19.694	14.325	35.719	62.54	5.5556	63.889	-14.782	7356	13238	4406	87	2	23	11
meyer-green-2019-03-16_1	22.389	21.872	23.03	31.543	78.746	1.5625	50	-2.2555	7384	16025	1993	150	1	32	31
nvidia-aud-2019-01-25_0	40.372	22.359	41.072	48.525	86.686	14.516	59.677	2.6964	16075	17052	2469	232	9	37	16
nvidia-aud-2019-04-18_1	47.758	16.859	48.096	57.64	85.794	29.63	33.333	-0.16478	4602	3382	762	27	8	9	10
nvidia-aud-2019-04-18_2	58.878	22.467	59.173	65.547	91.138	47.059	23.529	8.058	6674	3508	649	30	16	8	10
quarry-road-2019-02-28_0	22.766	20.37	23.984	48.189	66.565	8.3333	50	-15.607	2847	3061	1430	72	2	12	10
serra-street-2019-01-30_0	19.324	23.713	19.931	31.515	73.123	4.4248	61.062	-4.717	13557	29461	4983	261	5	69	39
stlc-111-2019-04-19_1	63.665	13.985	64.322	72.172	90.19	37.5	54.167	1.5864	4505	1737	490	41	9	13	2
stlc-111-2019-04-19_2	58.099	14.606	58.76	70.303	85.897	21.053	57.895	-1.9355	2552	1078	419	24	4	11	4
tressider-2019-03-16_2	17.967	23.091	18.833	35.903	67.776	10.909	43.636	-9.646	9570	17085	4550	231	6	24	25
tressider-2019-04-26_0	24.329	22.327	25.158	29.701	86.733	1.1628	55.814	1.2599	35559	84166	5439	992	3	144	111
tressider-2019-04-26_1	30.323	18.681	30.743	32.873	93.913	6.3927	40.183	3.5908	63293	129245	4102	808	14	88	117
tressider-2019-04-26_3	29.584	23.159	30.353	37.337	84.242	4.2735	52.991	0.89358	43553	73094	8147	897	10	124	100
COMBINED	32.326	20.121	32.946	39.877	85.194	11.604	52.102	0.47341	361501	545047	62828	5618	196	880	613

Sequence	IDF1	IDR	IDP	IDTP	IDFN	IDFP
cubberly-auditorium-2019-04-22_1	45.756	36.241	62.047	4795	8436	2933
discovery-walk-2019-02-28_0	33.008	27.339	41.645	3641	9677	5102
discovery-walk-2019-02-28_1	44.217	36.005	57.281	5216	9271	3890
food-trucks-2019-02-12_0	40.445	31.463	56.606	20167	43931	15460
gates-ai-lab-2019-04-17_0	26.384	21.388	34.426	5523	20300	10520
gates-basement-elevators-2019-01-17_0	51.83	40.189	72.965	4985	7419	1847
gates-foyer-2019-01-17_0	39.032	37.456	40.747	3816	6372	5549
gates-to-clark-2019-02-28_0	43.455	39.044	48.992	996	1555	1037
hewlett-class-2019-01-23_0	50.583	41.591	64.536	6877	9658	3779
hewlett-class-2019-01-23_1	75.202	70.722	80.287	3633	1504	892
huang-2-2019-01-25_1	39.224	33.394	47.521	2214	4416	2445
huang-intersection-2019-01-22_0	36.714	27.81	54.005	13040	33850	11106
indoor-coupa-cafe-2019-02-06_0	23.111	16.4	39.12	10758	54839	16742
lomita-serra-intersection-2019-01-30_0	37.588	29.528	51.7	6081	14513	5681
meyer-green-2019-03-16_1	32.837	22.995	57.406	5383	18026	3994
nvidia-aud-2019-01-25_0	37.805	29.484	52.669	9767	23360	8777
nvidia-aud-2019-04-18_1	60.458	50.539	75.224	4035	3949	1329
nvidia-aud-2019-04-18_2	68.118	58.554	81.415	5962	4220	1361
quarry-road-2019-02-28_0	41.885	36.104	49.871	2133	3775	2144
serra-street-2019-01-30_0	34.777	24.883	57.735	10704	32314	7836
stlc-111-2019-04-19_1	61.6	55.447	69.289	3461	2781	1534
stlc-111-2019-04-19_2	53.992	49.091	59.98	1782	1848	1189
tressider-2019-03-16_2	29.807	22.799	43.038	6077	20578	8043
tressider-2019-04-26_0	26.004	17.454	50.971	20897	98828	20101
tressider-2019-04-26_1	29.563	19.956	57.01	38422	154116	28973
tressider-2019-04-26_3	28.311	20.429	46.093	23830	92817	27870
COMBINED	33.691	24.731	52.835	224195	682353	200134

Sequence	Dets	IDs
cubberly-auditorium-2019-04-22_1	7728	86
discovery-walk-2019-02-28_0	8743	98
discovery-walk-2019-02-28_1	9106	114
food-trucks-2019-02-12_0	35627	234
gates-ai-lab-2019-04-17_0	16043	186
gates-basement-elevators-2019-01-17_0	6832	93
gates-foyer-2019-01-17_0	9365	81
gates-to-clark-2019-02-28_0	2033	40
hewlett-class-2019-01-23_0	10656	168
hewlett-class-2019-01-23_1	4525	54
huang-2-2019-01-25_1	4659	56
huang-intersection-2019-01-22_0	24146	187
indoor-coupa-cafe-2019-02-06_0	27500	277
lomita-serra-intersection-2019-01-30_0	11762	89
meyer-green-2019-03-16_1	9377	127
nvidia-aud-2019-01-25_0	18544	227
nvidia-aud-2019-04-18_1	5364	44
nvidia-aud-2019-04-18_2	7323	52
quarry-road-2019-02-28_0	4277	63
serra-street-2019-01-30_0	18540	185
stlc-111-2019-04-19_1	4995	58
stlc-111-2019-04-19_2	2971	40
tressider-2019-03-16_2	14120	153
tressider-2019-04-26_0	40998	853
tressider-2019-04-26_1	67395	557
tressider-2019-04-26_3	51700	625
COMBINED	424329	4747

Tracktor++_yw2

31.679%

0.95

28.152%

25.529%

Anonymous Submission

Details

Name	Tracktor++_yw2
Submission Date	2021-12-16 12:21:04+00:00
Abstract	The problem of tracking multiple objects in a video se- quence poses several challenging tasks. For tracking-by- detection, these include object re-identification, motion pre- diction and dealing with occlusions. We present a tracker (without bells and whistles) that accomplishes tracking without specifically targeting any of these tasks, in partic- ular, we perform no training or optimization on tracking data. To this end, we exploit the bounding box regression of an object detector to predict the position of an object in the next frame, thereby converting a detector into a Tracktor. We demonstrate the potential of Tracktor and provide a new state-of-the-art on three multi-object tracking benchmarks by extending it with a straightforward re-identification and camera motion compensation. We then perform an analysis on the performance and failure cases of several state-of-the-art tracking methods in comparison to our Tracktor. Surprisingly, none of the dedicated tracking methods are considerably better in deal- ing with complex tracking scenarios, namely, small and occluded objects or missing detections. However, our ap- proach tackles most of the easy tracking scenarios. There- fore, we motivate our approach as a new tracking paradigm and point out promising future research directions. Over- all, Tracktor yields superior tracking performance than any current tracking method and our analysis exposes remain- ing and unsolved tracking challenges to inspire future re- search directions.
Publication title	N/A	Publication authors:	N/A
Publication venue and year	N/A	Publication URL:	N/A
Code Language	N/A	Hardware	N/A
Code Website	N/A	Code URL	N/A

Visualisation

stlc-111-2019-04-19_1

Best 3 By MOTA ↓

Worst 3 By MOTA ↓

Per-sequence Results

Sequence	OSPA	OSPA_CARD	OSPA_LOC
cubberly-auditorium-2019-04-22_1	0.93257	0.77907	0.1535
discovery-walk-2019-02-28_0	0.97406	0.82895	0.14511
discovery-walk-2019-02-28_1	0.97548	0.85214	0.12334
food-trucks-2019-02-12_0	0.97301	0.85437	0.11864
gates-ai-lab-2019-04-17_0	0.99235	0.95883	0.033519
gates-basement-elevators-2019-01-17_0	0.89966	0.76048	0.13918
gates-foyer-2019-01-17_0	0.94795	0.87719	0.070756
gates-to-clark-2019-02-28_0	0.95435	0.86316	0.091195
hewlett-class-2019-01-23_0	0.9032	0.73256	0.17064
hewlett-class-2019-01-23_1	0.86498	0.76423	0.10075
huang-2-2019-01-25_1	0.96531	0.86232	0.10299
huang-intersection-2019-01-22_0	0.97501	0.79384	0.18117
indoor-coupa-cafe-2019-02-06_0	0.98812	0.90713	0.080989
lomita-serra-intersection-2019-01-30_0	0.97902	0.84937	0.12964
meyer-green-2019-03-16_1	0.98332	0.87621	0.10712
nvidia-aud-2019-01-25_0	0.97496	0.89199	0.082978
nvidia-aud-2019-04-18_1	0.90975	0.7033	0.20646
nvidia-aud-2019-04-18_2	0.8949	0.65657	0.23833
quarry-road-2019-02-28_0	0.98172	0.91011	0.071611
serra-street-2019-01-30_0	0.98819	0.8905	0.097687
stlc-111-2019-04-19_1	0.85748	0.70732	0.15017
stlc-111-2019-04-19_2	0.89348	0.72464	0.16884
tressider-2019-03-16_2	0.97617	0.80287	0.1733
tressider-2019-04-26_0	0.97683	0.84074	0.13609
tressider-2019-04-26_1	0.97851	0.82195	0.15656
tressider-2019-04-26_3	0.96854	0.82178	0.14675
COMBINED	0.95034	0.82045	0.1299

Sequence	HOTA	DetA	AssA	DetRe	DetPr	AssRe	AssPr	LocA
cubberly-auditorium-2019-04-22_1	33.595	38.06	30.034	43.146	65.414	32.727	69.855	79.167
discovery-walk-2019-02-28_0	28.074	31.86	26.41	34.064	71.568	28.019	75.749	79.735
discovery-walk-2019-02-28_1	29.703	35.122	25.411	38.222	69.607	26.892	72.92	79.431
food-trucks-2019-02-12_0	30.109	41.023	22.602	43.287	77.374	24.148	74.616	81.794
gates-ai-lab-2019-04-17_0	26.06	39.366	17.393	47.521	63.675	18.064	77.658	83.078
gates-basement-elevators-2019-01-17_0	39.268	45.114	34.503	49.724	72.811	37.688	77.092	81.85
gates-foyer-2019-01-17_0	38.937	53.142	28.618	63.4	68.95	29.893	81.001	83.235
gates-to-clark-2019-02-28_0	36.538	47.972	27.948	57.365	67.53	31.617	75.911	83.031
hewlett-class-2019-01-23_0	37.389	46.81	30.082	53.438	71.454	34.703	67.34	83.503
hewlett-class-2019-01-23_1	51.837	57.795	46.548	74.798	65.794	49.781	75.913	84.805
huang-2-2019-01-25_1	32.885	33.821	32.293	40.893	59.771	34.686	71.945	81.594
huang-intersection-2019-01-22_0	23.237	22.784	24.801	25.187	60.123	26.927	66.111	77.269
indoor-coupa-cafe-2019-02-06_0	19.857	29.514	13.598	31.669	70.729	14.106	75.393	80.186
lomita-serra-intersection-2019-01-30_0	29.812	25.723	35.127	28.464	62.883	36.919	69.903	78.546
meyer-green-2019-03-16_1	24.184	23.868	24.986	26.805	58.441	26.572	68.646	76.843
nvidia-aud-2019-01-25_0	27.181	35.397	21.942	39.652	64.454	23.837	71.207	77.333
nvidia-aud-2019-04-18_1	48.307	42.898	55.579	49.784	68.804	58.168	85.088	83.439
nvidia-aud-2019-04-18_2	41.65	44.58	39.286	49.425	69.146	42.816	68.083	78.935
quarry-road-2019-02-28_0	29.242	25.935	33.243	34.025	45.234	35.37	71.809	75.497
serra-street-2019-01-30_0	22.089	19.964	25.818	21.778	59.581	27.824	69.836	76.45
stlc-111-2019-04-19_1	53.775	57.883	50.16	64.324	76.13	53.09	79.517	83.599
stlc-111-2019-04-19_2	42.881	53.044	34.715	63.528	68.755	37.145	72.88	83.224
tressider-2019-03-16_2	24.186	21.063	29.166	22.391	65.235	31.634	68.328	76.871
tressider-2019-04-26_0	18.213	24.386	13.936	26.123	66.389	14.887	66.447	77.645
tressider-2019-04-26_1	20.909	26.256	16.959	27.543	73.204	18.404	67.749	79.975
tressider-2019-04-26_3	20.934	29.563	15.413	32.436	64.599	16.81	63.733	77.22
COMBINED	25.529	30.12	22.184	32.867	67.798	23.904	70.68	79.492

Sequence	MOTA	MOTP	MODA	CLR_Re	CLR_Pr	MTR	PTR	sMOTA	CLR_TP	CLR_FN	CLR_FP	IDSW	MT	PT	ML
cubberly-auditorium-2019-04-22_1	44.229	24.627	45.023	55.491	84.13	13.158	63.158	2.4043	7342	5889	1385	105	5	24	9
discovery-walk-2019-02-28_0	35.553	23.797	36.469	42.033	88.31	7.6923	53.846	3.5226	5598	7720	741	122	2	14	10
discovery-walk-2019-02-28_1	39.394	23.822	40.498	47.705	86.876	10.526	52.632	3.0534	6911	7576	1044	160	4	20	14
food-trucks-2019-02-12_0	50.097	21.405	50.722	53.334	95.332	20	55	8.1787	34186	29912	1674	401	12	33	15
gates-ai-lab-2019-04-17_0	34.849	18.471	35.495	55.063	73.781	29.167	37.5	-10.044	14219	11604	5053	167	7	9	8
gates-basement-elevators-2019-01-17_0	53.394	21.535	54.233	61.262	89.706	45	50	5.3247	7599	4805	872	104	18	20	2
gates-foyer-2019-01-17_0	60.12	19.298	60.679	76.315	82.995	38.095	52.381	-1.4682	7775	2413	1593	57	8	11	2
gates-to-clark-2019-02-28_0	49.275	19.203	50.686	67.817	79.834	53.846	23.077	-5.5192	1730	821	437	36	7	3	3
hewlett-class-2019-01-23_0	53.057	19.042	54.164	64.475	86.212	39.13	57.971	0.85942	10661	5874	1705	183	27	40	2
hewlett-class-2019-01-23_1	62.488	17.161	63.111	88.398	77.757	75.862	24.138	-10.74	4541	596	1299	32	22	7	0
huang-2-2019-01-25_1	30.226	21.31	30.95	49.683	72.619	21.053	52.632	-8.8693	3294	3336	1242	48	4	10	5
huang-intersection-2019-01-22_0	20.203	26.317	21.013	31.452	75.08	2.2989	55.172	-2.9724	14748	32142	4895	380	2	48	37
indoor-coupa-cafe-2019-02-06_0	31.7	22.527	32.293	38.534	86.061	5.3571	55.357	1.8463	25277	40320	4094	389	3	31	22
lomita-serra-intersection-2019-01-30_0	22.487	23.92	23.201	34.233	75.628	5.5556	61.111	-3.5577	7050	13544	2272	147	2	22	12
meyer-green-2019-03-16_1	19.283	26.065	20.441	33.154	72.283	1.5625	50	-5.2293	7761	15648	2976	271	1	32	31
nvidia-aud-2019-01-25_0	33.788	25.604	34.89	48.205	78.356	17.742	58.065	-2.075	15969	17158	4411	365	11	36	15
nvidia-aud-2019-04-18_1	41.809	18.409	42.247	57.302	79.193	37.037	18.519	-4.9448	4575	3409	1202	35	10	5	12
nvidia-aud-2019-04-18_2	54.93	25.412	55.47	63.475	88.802	38.235	35.294	7.586	6463	3719	815	55	13	12	9
quarry-road-2019-02-28_0	7.9892	28.021	8.6662	41.943	55.761	8.3333	50	-22.201	2478	3430	1966	40	2	12	10
serra-street-2019-01-30_0	14.268	26.61	16.017	26.284	71.909	5.3097	48.673	-5.0218	11307	31711	4417	752	6	55	52
stlc-111-2019-04-19_1	68.087	18.569	68.792	76.642	90.709	45.833	54.167	5.6764	4784	1458	490	44	11	13	0
stlc-111-2019-04-19_2	57.41	18.966	58.402	75.399	81.604	42.105	42.105	-3.6891	2737	893	617	36	8	8	3
tressider-2019-03-16_2	21.024	26.006	21.613	27.968	81.484	7.2727	30.909	0.32916	7455	19200	1694	157	4	17	34
tressider-2019-04-26_0	24.772	25.301	26.098	32.723	83.163	2.3256	62.403	0.32791	39178	80547	7932	1588	6	161	91
tressider-2019-04-26_1	30.208	22.632	30.941	34.283	91.117	7.7626	41.553	3.6833	66008	126530	6435	1412	17	91	111
tressider-2019-04-26_3	30.762	26.127	31.949	41.08	81.815	5.9829	58.12	0.41482	47919	68728	10651	1385	14	136	84
COMBINED	31.679	23.394	32.613	40.546	83.637	13.381	51.51	0.61841	367565	538983	71912	8471	226	870	593

Sequence	IDF1	IDR	IDP	IDTP	IDFN	IDFP
cubberly-auditorium-2019-04-22_1	45.277	37.571	56.961	4971	8260	3756
discovery-walk-2019-02-28_0	30.31	22.368	46.995	2979	10339	3360
discovery-walk-2019-02-28_1	36.467	28.246	51.439	4092	10395	3863
food-trucks-2019-02-12_0	32.578	25.402	45.404	16282	47816	19578
gates-ai-lab-2019-04-17_0	23.373	20.408	27.345	5270	20553	14002
gates-basement-elevators-2019-01-17_0	48.24	40.592	59.438	5035	7369	3436
gates-foyer-2019-01-17_0	42.054	40.361	43.894	4112	6076	5256
gates-to-clark-2019-02-28_0	36.329	33.595	39.548	857	1694	1310
hewlett-class-2019-01-23_0	43.846	38.319	51.237	6336	10199	6030
hewlett-class-2019-01-23_1	58.14	62.118	54.64	3191	1946	2649
huang-2-2019-01-25_1	40.659	34.238	50.044	2270	4360	2266
huang-intersection-2019-01-22_0	27.259	19.339	46.164	9068	37822	10575
indoor-coupa-cafe-2019-02-06_0	22.82	16.519	36.894	10836	54761	18535
lomita-serra-intersection-2019-01-30_0	34.136	24.794	54.774	5106	15488	4216
meyer-green-2019-03-16_1	28.466	20.761	45.264	4860	18549	5877
nvidia-aud-2019-01-25_0	31.401	25.36	41.222	8401	24726	11979
nvidia-aud-2019-04-18_1	53.005	45.679	63.13	3647	4337	2130
nvidia-aud-2019-04-18_2	50.183	43.027	60.195	4381	5801	2897
quarry-road-2019-02-28_0	39.084	34.242	45.522	2023	3885	2421
serra-street-2019-01-30_0	20.912	14.278	39.061	6142	36876	9582
stlc-111-2019-04-19_1	67.558	62.32	73.758	3890	2352	1384
stlc-111-2019-04-19_2	50.745	48.815	52.832	1772	1858	1582
tressider-2019-03-16_2	24.601	16.522	48.136	4404	22251	4745
tressider-2019-04-26_0	24.066	16.768	42.613	20075	99650	27035
tressider-2019-04-26_1	22.067	15.185	40.359	29237	163301	43206
tressider-2019-04-26_3	23.093	17.344	34.542	20231	96416	38339
COMBINED	28.152	20.9	43.112	189468	717080	250009

Sequence	Dets	IDs
cubberly-auditorium-2019-04-22_1	8727	172
discovery-walk-2019-02-28_0	6339	152
discovery-walk-2019-02-28_1	7955	257
food-trucks-2019-02-12_0	35860	412
gates-ai-lab-2019-04-17_0	19272	583
gates-basement-elevators-2019-01-17_0	8471	167
gates-foyer-2019-01-17_0	9368	171
gates-to-clark-2019-02-28_0	2167	95
hewlett-class-2019-01-23_0	12366	258
hewlett-class-2019-01-23_1	5840	123
huang-2-2019-01-25_1	4536	138
huang-intersection-2019-01-22_0	19643	422
indoor-coupa-cafe-2019-02-06_0	29371	603
lomita-serra-intersection-2019-01-30_0	9322	239
meyer-green-2019-03-16_1	10737	517
nvidia-aud-2019-01-25_0	20380	574
nvidia-aud-2019-04-18_1	5777	91
nvidia-aud-2019-04-18_2	7278	99
quarry-road-2019-02-28_0	4444	267
serra-street-2019-01-30_0	15724	1032
stlc-111-2019-04-19_1	5274	82
stlc-111-2019-04-19_2	3354	69
tressider-2019-03-16_2	9149	279
tressider-2019-04-26_0	47110	1620
tressider-2019-04-26_1	72443	1230
tressider-2019-04-26_3	58570	1313
COMBINED	439477	10965

Team inyoung

24.05%

0.892

27.312%

23.894%

Anonymous Submission

Details

Name	Team inyoung
Submission Date	2021-12-16 12:22:21+00:00
Abstract	Simple Online and Realtime Tracking (SORT) is a pragmatic approach to multiple object tracking with a focus on simple, effective algorithms. In this paper, we integrate appearance information to improve the performance of SORT. Due to this extension we are able to track objects through longer periods of occlusions, effectively reducing the number of identity switches. In spirit of the original framework we place much of the computational complexity into an offline pre-training stage where we learn a deep association metric on a largescale person re-identification dataset. During online application, we establish measurement-to-track associations using nearest neighbor queries in visual appearance space. Experimental evaluation shows that our extensions reduce the number of identity switches by 45%, achieving overall competitive performance at high frame rates.
Publication title	N/A	Publication authors:	N/A
Publication venue and year	N/A	Publication URL:	N/A
Code Language	N/A	Hardware	N/A
Code Website	N/A	Code URL	N/A

Visualisation

hewlett-class-2019-01-23_1

Best 3 By MOTA ↓

Worst 3 By MOTA ↓

Per-sequence Results

Sequence	OSPA	OSPA_CARD	OSPA_LOC
cubberly-auditorium-2019-04-22_1	0.89061	0.53659	0.35403
discovery-walk-2019-02-28_0	0.94504	0.61765	0.3274
discovery-walk-2019-02-28_1	0.93419	0.60417	0.33002
food-trucks-2019-02-12_0	0.92086	0.48718	0.43368
gates-ai-lab-2019-04-17_0	0.98671	0.92793	0.058785
gates-basement-elevators-2019-01-17_0	0.79897	0.5	0.29897
gates-foyer-2019-01-17_0	0.93668	0.82203	0.11464
gates-to-clark-2019-02-28_0	0.88956	0.65789	0.23167
hewlett-class-2019-01-23_0	0.68589	0.028169	0.65772
hewlett-class-2019-01-23_1	0.55658	0.23684	0.31974
huang-2-2019-01-25_1	0.90176	0.62	0.28176
huang-intersection-2019-01-22_0	0.92059	0.30952	0.61106
indoor-coupa-cafe-2019-02-06_0	0.98246	0.82222	0.16024
lomita-serra-intersection-2019-01-30_0	0.9259	0.42857	0.49733
meyer-green-2019-03-16_1	0.95979	0.56164	0.39815
nvidia-aud-2019-01-25_0	0.95727	0.75591	0.20137
nvidia-aud-2019-04-18_1	0.83802	0.37209	0.46592
nvidia-aud-2019-04-18_2	0.85974	0.54667	0.31308
quarry-road-2019-02-28_0	0.94922	0.72727	0.22195
serra-street-2019-01-30_0	0.92786	0.44059	0.48727
stlc-111-2019-04-19_1	0.80505	0.48936	0.31569
stlc-111-2019-04-19_2	0.86537	0.42424	0.44113
tressider-2019-03-16_2	0.92578	0.40217	0.5236
tressider-2019-04-26_0	0.95574	0.57774	0.378
tressider-2019-04-26_1	0.93637	0.4	0.53637
tressider-2019-04-26_3	0.93456	0.48571	0.44885
COMBINED	0.89195	0.53008	0.36186

Sequence	HOTA	DetA	AssA	DetRe	DetPr	AssRe	AssPr	LocA
cubberly-auditorium-2019-04-22_1	29.142	30.909	27.979	34.28	63.998	31.864	65.391	77.431
discovery-walk-2019-02-28_0	25.837	25.461	27.195	27.584	65.764	29.19	68.524	78.417
discovery-walk-2019-02-28_1	28.242	31.428	25.513	34.598	67.106	27.947	67.492	79.812
food-trucks-2019-02-12_0	27.399	29.674	25.497	30.889	75.823	27.479	62.636	80.525
gates-ai-lab-2019-04-17_0	23.913	32.017	18.023	44.938	48.694	19.287	67.351	81.511
gates-basement-elevators-2019-01-17_0	39.656	40.758	38.785	43.192	76.624	44.057	68.753	81.697
gates-foyer-2019-01-17_0	33.668	42.629	26.654	60.024	54.137	27.936	77.377	81.691
gates-to-clark-2019-02-28_0	38.352	41.831	35.559	49.079	65.756	37.518	76.711	81.152
hewlett-class-2019-01-23_0	38.899	46.933	32.47	50.068	76.954	41.315	57.15	81.662
hewlett-class-2019-01-23_1	61.661	66.284	57.416	71.268	79.158	61.664	78.93	82.843
huang-2-2019-01-25_1	32.915	32.789	33.846	39.025	58.911	36.837	68.601	78.948
huang-intersection-2019-01-22_0	22.79	18.469	28.357	20.968	52.35	31.467	58.85	76.373
indoor-coupa-cafe-2019-02-06_0	16.412	23.619	11.685	25.354	66.27	12.139	68.719	78.267
lomita-serra-intersection-2019-01-30_0	23.814	21.081	27.097	23.259	58.925	31.877	58.128	77.034
meyer-green-2019-03-16_1	21.508	16.332	29.667	17.852	57.036	32.105	68.823	77.464
nvidia-aud-2019-01-25_0	25.398	26.748	25.427	30.002	58.875	27.792	66.646	75.412
nvidia-aud-2019-04-18_1	43.212	38.998	48.327	43.842	68.823	51.241	78.164	81.023
nvidia-aud-2019-04-18_2	38.981	32.708	46.945	42.704	49.789	51.607	67.103	75.497
quarry-road-2019-02-28_0	26.24	20.866	33.07	28.917	38.435	35.874	69.845	76.246
serra-street-2019-01-30_0	22.779	14.945	36.388	16.574	52.043	39.221	71.775	76.153
stlc-111-2019-04-19_1	48.503	51.934	45.722	56.534	76.614	52.43	70.319	82.806
stlc-111-2019-04-19_2	36.525	47.171	28.312	54.182	69.596	36.854	62.039	81.978
tressider-2019-03-16_2	24.603	19.182	32.967	20.932	59.909	36.499	67.414	77.453
tressider-2019-04-26_0	15.745	17.613	14.387	18.435	66.967	15.736	62.67	77.241
tressider-2019-04-26_1	20.204	22.342	18.66	23.143	73.803	22.325	51.834	79.479
tressider-2019-04-26_3	18.749	22.359	16.195	23.848	65.046	18.331	58.463	76.655
COMBINED	23.894	24.256	23.956	26.414	64.531	26.86	62.89	78.653

Sequence	MOTA	MOTP	MODA	CLR_Re	CLR_Pr	MTR	PTR	sMOTA	CLR_TP	CLR_FN	CLR_FP	IDSW	MT	PT	ML
cubberly-auditorium-2019-04-22_1	29.59	25.755	30.3	41.932	78.284	13.158	52.632	-1.5426	5548	7683	1539	94	5	20	13
discovery-walk-2019-02-28_0	24.193	24.669	25.004	33.473	79.807	7.6923	53.846	-1.0232	4458	8860	1128	108	2	14	10
discovery-walk-2019-02-28_1	34.307	23.26	34.949	43.253	83.893	10.526	47.368	1.1145	6266	8221	1203	93	4	18	16
food-trucks-2019-02-12_0	36.923	22.37	37.302	39.02	95.784	15	38.333	6.6319	25011	39087	1101	243	9	23	28
gates-ai-lab-2019-04-17_0	11.467	20.014	12.156	52.221	56.586	29.167	37.5	-30.303	13485	12338	10346	178	7	9	8
gates-basement-elevators-2019-01-17_0	48.686	21.166	49.355	52.862	93.779	37.5	52.5	7.0128	6557	5847	435	83	15	21	4
gates-foyer-2019-01-17_0	34.03	20.752	34.433	72.654	65.528	28.571	61.905	-23.547	7402	2786	3894	41	6	13	2
gates-to-clark-2019-02-28_0	39.867	20.917	41.082	57.86	77.521	38.462	38.462	-5.8903	1476	1075	428	31	5	5	3
hewlett-class-2019-01-23_0	55.392	20.408	56.22	60.641	93.205	31.884	66.667	7.1259	10027	6508	731	137	22	46	1
hewlett-class-2019-01-23_1	83.317	20.033	83.609	86.821	96.432	68.966	27.586	13.889	4460	677	165	15	20	8	1
huang-2-2019-01-25_1	27.647	23.896	28.175	47.21	71.266	15.789	52.632	-8.2816	3130	3500	1262	35	3	10	6
huang-intersection-2019-01-22_0	13.434	27.152	13.775	26.914	67.196	2.2989	43.678	-6.1727	12620	34270	6161	160	2	38	47
indoor-coupa-cafe-2019-02-06_0	23.358	24.863	23.997	31.128	81.36	1.7857	39.286	-0.030891	20419	45178	4678	419	1	22	33
lomita-serra-intersection-2019-01-30_0	19.054	26.39	19.214	29.343	74.339	5.5556	38.889	-2.5456	6043	14551	2086	33	2	14	20
meyer-green-2019-03-16_1	11.94	25.709	12.358	21.829	69.742	3.125	28.125	-4.2773	5110	18299	2217	98	2	18	44
nvidia-aud-2019-01-25_0	20.877	27.341	21.623	36.291	71.216	12.903	38.71	-5.491	12022	21105	4859	247	8	24	30
nvidia-aud-2019-04-18_1	36.899	20.453	37.224	50.463	79.217	29.63	25.926	-3.2433	4029	3955	1057	26	8	7	12
nvidia-aud-2019-04-18_2	20.674	27.862	21.224	53.496	62.373	23.529	47.059	-17.917	5447	4735	3286	56	8	16	10
quarry-road-2019-02-28_0	-3.8253	27.019	-3.4022	35.917	47.739	4.1667	41.667	-30.038	2122	3786	2323	25	1	10	13
serra-street-2019-01-30_0	8.5197	27.012	8.8568	20.352	63.905	3.5398	39.823	-6.3348	8755	34263	4945	145	4	45	64
stlc-111-2019-04-19_1	61.679	19.376	62.32	68.055	92.228	33.333	58.333	6.81	4248	1994	358	40	8	14	2
stlc-111-2019-04-19_2	53.113	20.639	53.774	65.813	84.536	21.053	47.368	0.88359	2389	1241	437	24	4	9	6
tressider-2019-03-16_2	14.826	24.896	15.235	25.087	71.803	7.2727	27.273	-4.015	6687	19968	2626	109	4	15	36
tressider-2019-04-26_0	18.024	25.802	18.776	23.152	84.104	1.5504	38.76	0.84538	27719	92006	5239	901	4	100	154
tressider-2019-04-26_1	27.079	23.755	27.578	29.468	93.973	5.9361	35.616	4.6107	56738	135800	3639	962	13	78	128
tressider-2019-04-26_3	22.548	26.583	23.369	30.016	81.869	3.8462	43.162	0.51141	35013	81634	7754	957	9	101	124
COMBINED	24.05	24.261	24.63	32.782	80.086	10.42	41.326	-0.77867	297181	609367	73897	5260	176	698	815

Sequence	IDF1	IDR	IDP	IDTP	IDFN	IDFP
cubberly-auditorium-2019-04-22_1	36.923	28.35	52.928	3751	9480	3336
discovery-walk-2019-02-28_0	29.687	21.069	50.233	2806	10512	2780
discovery-walk-2019-02-28_1	37.247	28.225	54.746	4089	10398	3380
food-trucks-2019-02-12_0	30.848	21.707	53.286	13914	50184	12198
gates-ai-lab-2019-04-17_0	22.838	21.957	23.793	5670	20153	18161
gates-basement-elevators-2019-01-17_0	51.495	40.261	71.424	4994	7410	1998
gates-foyer-2019-01-17_0	37.675	39.723	35.827	4047	6141	7249
gates-to-clark-2019-02-28_0	43.861	38.299	51.313	977	1574	927
hewlett-class-2019-01-23_0	47.712	39.377	60.522	6511	10024	4247
hewlett-class-2019-01-23_1	79.369	75.414	83.762	3874	1263	751
huang-2-2019-01-25_1	41.535	34.525	52.117	2289	4341	2103
huang-intersection-2019-01-22_0	29.508	20.663	51.589	9689	37201	9092
indoor-coupa-cafe-2019-02-06_0	19.651	13.584	35.506	8911	56686	16186
lomita-serra-intersection-2019-01-30_0	32.065	22.361	56.649	4605	15989	3524
meyer-green-2019-03-16_1	21.896	14.375	45.926	3365	20044	3962
nvidia-aud-2019-01-25_0	28.211	21.294	41.787	7054	26073	9827
nvidia-aud-2019-04-18_1	48.6	39.78	62.446	3176	4808	1910
nvidia-aud-2019-04-18_2	48.681	45.217	52.72	4604	5578	4129
quarry-road-2019-02-28_0	35.082	30.738	40.855	1816	4092	2629
serra-street-2019-01-30_0	25.678	16.928	53.153	7282	35736	6418
stlc-111-2019-04-19_1	62.223	54.069	73.274	3375	2867	1231
stlc-111-2019-04-19_2	42.069	37.41	48.054	1358	2272	1468
tressider-2019-03-16_2	24.772	16.714	47.836	4455	22200	4858
tressider-2019-04-26_0	21.252	13.551	49.226	16224	103501	16734
tressider-2019-04-26_1	21.989	14.442	46.056	27807	164731	32570
tressider-2019-04-26_3	22.366	15.283	41.684	17827	98820	24940
COMBINED	27.312	19.246	47.017	174470	732078	196608

Sequence	Dets	IDs
cubberly-auditorium-2019-04-22_1	7087	82
discovery-walk-2019-02-28_0	5586	68
discovery-walk-2019-02-28_1	7469	96
food-trucks-2019-02-12_0	26112	117
gates-ai-lab-2019-04-17_0	23831	333
gates-basement-elevators-2019-01-17_0	6992	80
gates-foyer-2019-01-17_0	11296	118
gates-to-clark-2019-02-28_0	1904	38
hewlett-class-2019-01-23_0	10758	71
hewlett-class-2019-01-23_1	4625	38
huang-2-2019-01-25_1	4392	50
huang-intersection-2019-01-22_0	18781	126
indoor-coupa-cafe-2019-02-06_0	25097	315
lomita-serra-intersection-2019-01-30_0	8129	63
meyer-green-2019-03-16_1	7327	146
nvidia-aud-2019-01-25_0	16881	254
nvidia-aud-2019-04-18_1	5086	43
nvidia-aud-2019-04-18_2	8733	75
quarry-road-2019-02-28_0	4445	88
serra-street-2019-01-30_0	13700	202
stlc-111-2019-04-19_1	4606	47
stlc-111-2019-04-19_2	2826	33
tressider-2019-03-16_2	9313	92
tressider-2019-04-26_0	32958	611
tressider-2019-04-26_1	60377	365
tressider-2019-04-26_3	42767	455
COMBINED	371078	4006

DeepSORT

23.878%

0.892

28.031%

24.379%

N. Wojke, A. Bewley and D. Paulus Simple Online and Realtime Tracking with a Deep Association Metric in ICIP, 2017

Details

Name	DeepSORT
Submission Date	2021-12-16 12:23:03+00:00
Abstract	Simple Online and Realtime Tracking (SORT) is a pragmatic approach to multiple object tracking with a focus on simple, effective algorithms. In this paper, we integrate appearance information to improve the performance of SORT. Due to this extension we are able to track objects through longer periods of occlusions, effectively reducing the number of identity switches. In spirit of the original framework we place much of the computational complexity into an offline pre-training stage where we learn a deep association metric on a large-scale person re-identification dataset. During online application, we establish measurement-to-track associations using nearest neighbor queries in visual appearance space. Experimental evaluation shows that our extensions reduce the number of identity switches by 45%, achieving overall competitive performance at high frame rates.
Publication title	Simple Online and Realtime Tracking with a Deep Association Metric	Publication authors:	N. Wojke, A. Bewley and D. Paulus
Publication venue and year	ICIP, 2017	Publication URL:	https://arxiv.org/abs/1703.07402
Code Language	N/A	Hardware	N/A
Code Website	N/A	Code URL	https://github.com/nwojke/deep_sort

Visualisation

hewlett-class-2019-01-23_1

Best 3 By MOTA ↓

Worst 3 By MOTA ↓

Per-sequence Results

Sequence	OSPA	OSPA_CARD	OSPA_LOC
cubberly-auditorium-2019-04-22_1	0.8922	0.54217	0.35003
discovery-walk-2019-02-28_0	0.9458	0.61765	0.32815
discovery-walk-2019-02-28_1	0.92997	0.60417	0.3258
food-trucks-2019-02-12_0	0.92238	0.48718	0.4352
gates-ai-lab-2019-04-17_0	0.98748	0.92878	0.058692
gates-basement-elevators-2019-01-17_0	0.80535	0.5	0.30535
gates-foyer-2019-01-17_0	0.93646	0.82203	0.11443
gates-to-clark-2019-02-28_0	0.89382	0.66667	0.22716
hewlett-class-2019-01-23_0	0.67497	0	0.67497
hewlett-class-2019-01-23_1	0.55005	0.25641	0.29364
huang-2-2019-01-25_1	0.8984	0.62	0.2784
huang-intersection-2019-01-22_0	0.91915	0.28689	0.63227
indoor-coupa-cafe-2019-02-06_0	0.98242	0.82222	0.1602
lomita-serra-intersection-2019-01-30_0	0.92633	0.42857	0.49776
meyer-green-2019-03-16_1	0.96002	0.56164	0.39838
nvidia-aud-2019-01-25_0	0.95749	0.75591	0.20158
nvidia-aud-2019-04-18_1	0.84261	0.38636	0.45625
nvidia-aud-2019-04-18_2	0.86544	0.54054	0.3249
quarry-road-2019-02-28_0	0.95132	0.73626	0.21506
serra-street-2019-01-30_0	0.92958	0.44608	0.4835
stlc-111-2019-04-19_1	0.80627	0.48936	0.31691
stlc-111-2019-04-19_2	0.85394	0.44118	0.41277
tressider-2019-03-16_2	0.93125	0.43299	0.49826
tressider-2019-04-26_0	0.95489	0.57843	0.37646
tressider-2019-04-26_1	0.93228	0.42368	0.50859
tressider-2019-04-26_3	0.93538	0.50424	0.43115
COMBINED	0.89174	0.53382	0.35792

Sequence	HOTA	DetA	AssA	DetRe	DetPr	AssRe	AssPr	LocA
cubberly-auditorium-2019-04-22_1	29.223	30.676	28.324	34.053	63.638	32.224	65.287	77.245
discovery-walk-2019-02-28_0	25.664	25.293	27.095	27.436	65.435	29.071	68.346	78.308
discovery-walk-2019-02-28_1	30.752	31.34	30.297	34.521	66.824	32.012	72.578	79.618
food-trucks-2019-02-12_0	27.357	29.56	25.548	30.768	75.533	27.708	62.828	80.281
gates-ai-lab-2019-04-17_0	22.845	31.964	16.487	44.864	48.612	17.919	66.897	81.384
gates-basement-elevators-2019-01-17_0	38.871	40.618	37.404	43.05	76.383	43.014	67.281	81.483
gates-foyer-2019-01-17_0	33.99	42.574	27.196	59.876	54.036	28.475	77.947	81.491
gates-to-clark-2019-02-28_0	38.06	41.274	35.445	48.565	65.308	37.083	77.868	81.156
hewlett-class-2019-01-23_0	41.506	46.801	37.222	49.967	76.749	46.038	59.289	81.557
hewlett-class-2019-01-23_1	62.309	66.054	58.828	71.043	78.976	62.433	80.43	82.741
huang-2-2019-01-25_1	33.756	32.752	35.501	38.984	58.875	38.706	68.911	78.874
huang-intersection-2019-01-22_0	22.49	18.266	27.942	20.765	51.715	30.987	57.441	75.964
indoor-coupa-cafe-2019-02-06_0	16.425	23.57	11.728	25.303	66.173	12.223	67.71	78.2
lomita-serra-intersection-2019-01-30_0	23.681	20.994	26.895	23.169	58.675	31.637	57.845	76.855
meyer-green-2019-03-16_1	21.469	16.315	29.568	17.828	57.059	31.998	68.902	77.391
nvidia-aud-2019-01-25_0	25.449	26.753	25.452	30.005	58.801	27.966	67.78	75.324
nvidia-aud-2019-04-18_1	43.134	38.827	48.367	43.664	68.651	51.283	78.214	80.88
nvidia-aud-2019-04-18_2	36.906	32.537	42.485	42.522	49.51	47.245	65.141	75.356
quarry-road-2019-02-28_0	26.137	20.635	33.189	28.599	38.149	35.995	69.931	76.094
serra-street-2019-01-30_0	22.641	14.828	36.218	16.456	51.684	39.055	71.604	76.013
stlc-111-2019-04-19_1	48.804	51.957	46.303	56.455	76.574	52.116	72.862	82.637
stlc-111-2019-04-19_2	38.284	47.076	31.156	54.074	69.508	37.322	67.803	81.841
tressider-2019-03-16_2	24.375	19.106	32.392	20.855	59.715	35.754	66.707	77.333
tressider-2019-04-26_0	15.895	17.561	14.699	18.385	66.779	15.94	63.192	77.142
tressider-2019-04-26_1	22.596	22.175	23.385	22.987	73.392	26.102	60.894	79.301
tressider-2019-04-26_3	18.878	22.262	16.487	23.747	64.899	18.479	60.309	76.572
COMBINED	24.379	24.147	25.047	26.304	64.287	27.728	64.987	78.505

Sequence	MOTA	MOTP	MODA	CLR_Re	CLR_Pr	MTR	PTR	sMOTA	CLR_TP	CLR_FN	CLR_FP	IDSW	MT	PT	ML
cubberly-auditorium-2019-04-22_1	29.204	25.998	29.899	41.705	77.938	13.158	52.632	-1.6585	5518	7713	1562	92	5	20	13
discovery-walk-2019-02-28_0	23.48	24.635	24.313	33.121	78.994	7.6923	53.846	-1.4818	4411	8907	1173	111	2	14	10
discovery-walk-2019-02-28_1	34.12	23.436	34.721	43.19	83.605	10.526	47.368	1.0518	6257	8230	1227	87	4	18	16
food-trucks-2019-02-12_0	36.831	22.705	37.206	38.97	95.668	15	38.333	6.7094	24979	39119	1131	240	9	23	28
gates-ai-lab-2019-04-17_0	12.005	20.206	12.702	52.496	56.882	29.167	37.5	-29.884	13556	12267	10276	180	7	9	8
gates-basement-elevators-2019-01-17_0	48.541	21.337	49.266	52.814	93.706	37.5	52.5	6.9959	6551	5853	440	90	15	21	4
gates-foyer-2019-01-17_0	34.138	21.025	34.521	72.664	65.577	28.571	61.905	-23.248	7403	2785	3886	39	6	13	2
gates-to-clark-2019-02-28_0	38.652	20.809	39.788	57.076	76.753	30.769	46.154	-6.5474	1456	1095	441	29	4	6	3
hewlett-class-2019-01-23_0	55.132	20.481	55.96	60.532	92.977	31.884	66.667	6.9968	10009	6526	756	137	22	46	1
hewlett-class-2019-01-23_1	83.161	20.164	83.414	86.685	96.364	68.966	27.586	13.955	4453	684	168	13	20	8	1
huang-2-2019-01-25_1	27.33	23.943	27.843	47.029	71.025	15.789	52.632	-8.4382	3118	3512	1272	34	3	10	6
huang-intersection-2019-01-22_0	12.821	27.526	13.21	26.682	66.449	2.2989	43.678	-6.5158	12511	34379	6317	182	2	38	47
indoor-coupa-cafe-2019-02-06_0	23.275	24.942	23.911	31.075	81.266	1.7857	39.286	-0.04853	20384	45213	4699	417	1	22	33
lomita-serra-intersection-2019-01-30_0	18.981	26.629	19.141	29.314	74.238	5.5556	38.889	-2.527	6037	14557	2095	33	2	14	20
meyer-green-2019-03-16_1	11.812	25.7	12.243	21.744	69.593	3.125	28.125	-4.3439	5090	18319	2224	101	2	18	44
nvidia-aud-2019-01-25_0	20.711	27.483	21.409	36.218	70.977	12.903	38.71	-5.5532	11998	21129	4906	231	8	24	30
nvidia-aud-2019-04-18_1	35.421	20.257	35.721	49.662	78.082	29.63	25.926	-4.1811	3965	4019	1113	24	8	7	12
nvidia-aud-2019-04-18_2	19.721	27.837	20.301	53.094	61.818	23.529	47.059	-18.593	5406	4776	3339	59	8	16	10
quarry-road-2019-02-28_0	-4.5362	27.15	-4.1469	35.41	47.234	4.1667	41.667	-30.332	2092	3816	2337	23	1	10	13
serra-street-2019-01-30_0	7.9967	27.065	8.3337	20.087	63.087	3.5398	38.938	-6.6537	8641	34377	5056	145	4	44	65
stlc-111-2019-04-19_1	61.599	19.649	62.224	67.975	92.199	33.333	58.333	6.9799	4243	1999	359	39	8	14	2
stlc-111-2019-04-19_2	53.278	20.818	53.939	65.868	84.667	21.053	47.368	1.1231	2391	1239	433	24	4	9	6
tressider-2019-03-16_2	14.207	24.787	14.62	24.772	70.931	7.2727	25.455	-4.4243	6603	20052	2706	110	4	14	37
tressider-2019-04-26_0	17.983	25.921	18.722	23.126	84.002	1.5504	39.535	0.85107	27688	92037	5273	885	4	102	152
tressider-2019-04-26_1	27.033	23.966	27.493	29.407	93.889	5.9361	35.616	4.674	56619	135919	3685	885	13	78	128
tressider-2019-04-26_3	22.439	26.706	23.243	29.917	81.76	3.8462	43.59	0.5123	34897	81750	7785	937	9	102	123
COMBINED	23.878	24.416	24.446	32.682	79.873	10.361	41.445	-0.82363	296276	610272	74659	5147	175	700	814

Sequence	IDF1	IDR	IDP	IDTP	IDFN	IDFP
cubberly-auditorium-2019-04-22_1	37.418	28.72	53.672	3800	9431	3280
discovery-walk-2019-02-28_0	29.5	20.934	49.928	2788	10530	2796
discovery-walk-2019-02-28_1	40.335	30.586	59.206	4431	10056	3053
food-trucks-2019-02-12_0	30.722	21.618	53.072	13857	50241	12253
gates-ai-lab-2019-04-17_0	21.315	20.493	22.205	5292	20531	18540
gates-basement-elevators-2019-01-17_0	49.569	38.754	68.76	4807	7597	2184
gates-foyer-2019-01-17_0	38.469	40.548	36.593	4131	6057	7158
gates-to-clark-2019-02-28_0	43.21	37.672	50.659	961	1590	936
hewlett-class-2019-01-23_0	50.308	41.53	63.79	6867	9668	3898
hewlett-class-2019-01-23_1	80.508	76.465	85.003	3928	1209	693
huang-2-2019-01-25_1	42.559	35.37	53.417	2345	4285	2045
huang-intersection-2019-01-22_0	28.838	20.209	50.329	9476	37414	9352
indoor-coupa-cafe-2019-02-06_0	19.711	13.624	35.63	8937	56660	16146
lomita-serra-intersection-2019-01-30_0	32.041	22.346	56.591	4602	15992	3530
meyer-green-2019-03-16_1	21.892	14.366	45.98	3363	20046	3951
nvidia-aud-2019-01-25_0	28.462	21.493	42.12	7120	26007	9784
nvidia-aud-2019-04-18_1	47.864	39.153	61.56	3126	4858	1952
nvidia-aud-2019-04-18_2	44.17	41.053	47.799	4180	6002	4565
quarry-road-2019-02-28_0	34.788	30.433	40.596	1798	4110	2631
serra-street-2019-01-30_0	25.298	16.677	52.376	7174	35844	6523
stlc-111-2019-04-19_1	62.007	53.861	73.055	3362	2880	1240
stlc-111-2019-04-19_2	46.886	41.68	53.576	1513	2117	1311
tressider-2019-03-16_2	23.824	16.072	46.02	4284	22371	5025
tressider-2019-04-26_0	21.631	13.793	50.102	16514	103211	16447
tressider-2019-04-26_1	25.686	16.865	53.847	32472	160066	27832
tressider-2019-04-26_3	22.494	15.363	41.985	17920	98727	24762
COMBINED	28.031	19.751	48.269	179048	727500	191887

Sequence	Dets	IDs
cubberly-auditorium-2019-04-22_1	7080	83
discovery-walk-2019-02-28_0	5584	68
discovery-walk-2019-02-28_1	7484	96
food-trucks-2019-02-12_0	26110	117
gates-ai-lab-2019-04-17_0	23832	337
gates-basement-elevators-2019-01-17_0	6991	80
gates-foyer-2019-01-17_0	11289	118
gates-to-clark-2019-02-28_0	1897	39
hewlett-class-2019-01-23_0	10765	69
hewlett-class-2019-01-23_1	4621	39
huang-2-2019-01-25_1	4390	50
huang-intersection-2019-01-22_0	18828	122
indoor-coupa-cafe-2019-02-06_0	25083	315
lomita-serra-intersection-2019-01-30_0	8132	63
meyer-green-2019-03-16_1	7314	146
nvidia-aud-2019-01-25_0	16904	254
nvidia-aud-2019-04-18_1	5078	44
nvidia-aud-2019-04-18_2	8745	74
quarry-road-2019-02-28_0	4429	91
serra-street-2019-01-30_0	13697	204
stlc-111-2019-04-19_1	4602	47
stlc-111-2019-04-19_2	2824	34
tressider-2019-03-16_2	9309	97
tressider-2019-04-26_0	32961	612
tressider-2019-04-26_1	60304	380
tressider-2019-04-26_3	42682	472
COMBINED	370935	4051

JRMOT2D

23.132%

0.945

22.168%

21.09%

A. Shenoi, M. Patel, J. Gwak, P. Goebel, A. Sadeghian, H. Rezatofighi, R. Martín-Martín and S. Savar JRMOT: A Real-Time 3D Multi-Object Tracker and a New Large-Scale Dataset in IROS, 2020

Details

Name	JRMOT2D
Submission Date	2021-12-16 12:24:07+00:00
Abstract	A robotic agent navigating autonomously needs to perceive and track the motion of objects and other agents in its surroundings to be able to plan and execute robust and safe trajectories. To be useful for planning and execution, the motion should be perceived in 3D Cartesian space. However, most recent multi-object tracking (MOT) research have focused on tracking people and moving objects in 2D RGB video sequences. This is because searching in 3D is costly, there is a lack of annotated datasets with 3D labels of moving agents and a relative scarcity of data with 3D sensor modalities. In this work we present JRMOT, a novel 3D MOT system that integrates information from RGB images and 3D point clouds to achieve real-time, state-of-the-art tracking performance. Our system incorporates recent advancements in neural network based re-identification as well as 2D and 3D detectors and descriptors. We integrate them into a joint probabilistic data-association framework within a multi-modal recursive Kalman architecture to achieve online, real-time 3D MOT. As part of our work, we release the JRDB dataset, a novel large scale 2D+3D dataset and benchmark annotated with over 2 million boxes and 3500 time consistent 2D+3D trajectories across 54 indoor and outdoor scenes. This dataset, that we use to train and test our model, contains over 60 minutes of data including 360 degree cylindrical RGB video and 3D pointclouds, in social settings. The presented 3D MOT system demonstrates state-ofthe-art performance against competing methods on the popular 2D tracking KITTI benchmark and serves as a competitive 3D tracking baseline for our dataset and benchmark. Moreover, our tests on our social robot JackRabbot indicate that the system is capable of tracking multiple pedestrians fast and reliably.
Publication title	JRMOT: A Real-Time 3D Multi-Object Tracker and a New Large-Scale Dataset	Publication authors:	A. Shenoi, M. Patel, J. Gwak, P. Goebel, A. Sadeghian, H. Rezatofighi, R. Martín-Martín and S. Savar
Publication venue and year	IROS, 2020	Publication URL:	https://arxiv.org/pdf/2002.08397.pdf
Code Language	N/A	Hardware	N/A
Code Website	N/A	Code URL	https://sites.google.com/view/jrmot

Visualisation

hewlett-class-2019-01-23_1

Best 3 By MOTA ↓

Worst 3 By MOTA ↓

Per-sequence Results

Sequence	OSPA	OSPA_CARD	OSPA_LOC
cubberly-auditorium-2019-04-22_1	0.93812	0.73611	0.20201
discovery-walk-2019-02-28_0	0.97554	0.78151	0.19403
discovery-walk-2019-02-28_1	0.96631	0.74324	0.22307
food-trucks-2019-02-12_0	0.97441	0.81424	0.16016
gates-ai-lab-2019-04-17_0	0.99411	0.96256	0.031553
gates-basement-elevators-2019-01-17_0	0.8798	0.70588	0.17391
gates-foyer-2019-01-17_0	0.96527	0.89806	0.067209
gates-to-clark-2019-02-28_0	0.91575	0.75	0.16575
hewlett-class-2019-01-23_0	0.87638	0.67143	0.20495
hewlett-class-2019-01-23_1	0.76138	0.56716	0.19421
huang-2-2019-01-25_1	0.96095	0.81553	0.14541
huang-intersection-2019-01-22_0	0.96821	0.58768	0.38053
indoor-coupa-cafe-2019-02-06_0	0.99289	0.9085	0.084394
lomita-serra-intersection-2019-01-30_0	0.96371	0.66972	0.29398
meyer-green-2019-03-16_1	0.97357	0.66316	0.31041
nvidia-aud-2019-01-25_0	0.98289	0.87525	0.10764
nvidia-aud-2019-04-18_1	0.91297	0.6625	0.25047
nvidia-aud-2019-04-18_2	0.94352	0.78481	0.15871
quarry-road-2019-02-28_0	0.96831	0.80488	0.16343
serra-street-2019-01-30_0	0.97041	0.68169	0.28872
stlc-111-2019-04-19_1	0.90781	0.72414	0.18367
stlc-111-2019-04-19_2	0.87252	0.62	0.25252
tressider-2019-03-16_2	0.96263	0.61268	0.34995
tressider-2019-04-26_0	0.98154	0.77701	0.20454
tressider-2019-04-26_1	0.9792	0.7959	0.1833
tressider-2019-04-26_3	0.97825	0.81159	0.16666
COMBINED	0.94486	0.74712	0.19774

Sequence	HOTA	DetA	AssA	DetRe	DetPr	AssRe	AssPr	LocA
cubberly-auditorium-2019-04-22_1	27.475	30.27	25.831	33.352	65.541	27.612	74.29	78.43
discovery-walk-2019-02-28_0	20.407	23.792	17.829	25.643	66.532	18.455	79.349	79.145
discovery-walk-2019-02-28_1	25.033	30.399	20.767	32.894	70.099	21.939	73.454	80.846
food-trucks-2019-02-12_0	24.641	30.033	20.666	31.057	77.354	21.713	78.639	80.828
gates-ai-lab-2019-04-17_0	22.364	32.709	15.447	44.106	51.954	15.9	80.337	82.736
gates-basement-elevators-2019-01-17_0	38.115	40.846	35.895	42.903	78.897	37.409	82.763	82.626
gates-foyer-2019-01-17_0	29.316	43.077	20.091	59.45	55.792	21.201	79.72	82.46
gates-to-clark-2019-02-28_0	47.205	41.964	53.725	48.662	69.003	59.508	80.021	83.95
hewlett-class-2019-01-23_0	39.091	47.871	32.177	50.718	78.944	34.05	81.705	82.665
hewlett-class-2019-01-23_1	60.71	67.792	54.444	72.714	80.833	56.95	83.71	83.952
huang-2-2019-01-25_1	31.601	32.694	31.911	37.998	62.204	36.455	75.665	80.287
huang-intersection-2019-01-22_0	13.628	14.103	13.226	14.832	62.162	13.718	72.756	76.477
indoor-coupa-cafe-2019-02-06_0	14.258	22.442	9.4223	23.943	67.391	9.623	76.242	78.858
lomita-serra-intersection-2019-01-30_0	23.442	20.432	27.133	22.599	58.46	27.802	77.786	77.445
meyer-green-2019-03-16_1	19.685	15.286	26.287	16.296	61.929	27.748	74.424	78.418
nvidia-aud-2019-01-25_0	19.592	25.662	15.904	28.507	60.48	17.317	73.049	76.373
nvidia-aud-2019-04-18_1	42.296	39.005	46.395	44.089	68.9	48.72	83.688	81.749
nvidia-aud-2019-04-18_2	32.845	33.463	33.351	42.444	52.119	35.542	75.492	75.778
quarry-road-2019-02-28_0	25.277	21.351	30.141	27.617	43.418	32.53	72.689	77.295
serra-street-2019-01-30_0	15.447	13.14	18.318	14.09	56.855	19.314	74.117	76.814
stlc-111-2019-04-19_1	43.558	51.784	37.093	56.75	76.957	38.753	80.219	83.865
stlc-111-2019-04-19_2	41.24	48.224	35.301	55.299	71.411	37.73	78.989	83.538
tressider-2019-03-16_2	19.893	18.523	21.73	19.471	67.228	22.562	76.886	77.744
tressider-2019-04-26_0	12.97	16.279	10.716	17.049	66.425	11.29	68.643	77.633
tressider-2019-04-26_1	18.432	22.205	15.64	22.925	74.861	16.128	79.155	79.861
tressider-2019-04-26_3	15.399	20.922	11.854	22.273	65.02	12.455	73.007	76.989
COMBINED	21.09	23.447	19.427	25.209	66.851	20.443	76.874	79.326

Sequence	MOTA	MOTP	MODA	CLR_Re	CLR_Pr	MTR	PTR	sMOTA	CLR_TP	CLR_FN	CLR_FP	IDSW	MT	PT	ML
cubberly-auditorium-2019-04-22_1	29.741	24.837	30.844	40.866	80.306	13.158	44.737	-0.97536	5407	7824	1326	146	5	17	16
discovery-walk-2019-02-28_0	21.73	23.582	22.443	30.493	79.116	7.6923	42.308	-1.5719	4061	9257	1072	95	2	11	13
discovery-walk-2019-02-28_1	32.712	21.874	33.52	40.222	85.716	10.526	42.105	1.288	5827	8660	971	117	4	16	18
food-trucks-2019-02-12_0	36.586	22.388	37.101	38.625	96.204	11.667	43.333	6.6085	24758	39340	977	330	7	26	27
gates-ai-lab-2019-04-17_0	15.471	18.604	16.613	50.753	59.785	25	37.5	-25.84	13106	12717	8816	295	6	9	9
gates-basement-elevators-2019-01-17_0	48.154	20.231	48.992	51.685	95.048	32.5	57.5	6.9255	6411	5993	334	104	13	23	4
gates-foyer-2019-01-17_0	35.08	19.774	36.042	71.3	66.912	33.333	57.143	-22.12	7264	2924	3592	98	7	12	2
gates-to-clark-2019-02-28_0	38.808	17.146	40.18	55.351	78.488	30.769	46.154	-7.0522	1412	1139	387	35	4	6	3
hewlett-class-2019-01-23_0	54.448	19.499	55.44	59.843	93.147	30.435	66.667	6.2744	9895	6640	728	164	21	46	2
hewlett-class-2019-01-23_1	82.422	18.461	83.142	86.549	96.213	68.966	24.138	11.85	4446	691	175	37	20	7	2
huang-2-2019-01-25_1	29.442	22.599	30.317	45.701	74.815	15.789	57.895	-5.9312	3030	3600	1020	58	3	11	5
huang-intersection-2019-01-22_0	13.171	26.969	13.504	18.682	78.298	0	24.138	-0.47233	8760	38130	2428	156	0	21	66
indoor-coupa-cafe-2019-02-06_0	21.734	24.298	22.728	29.128	81.987	1.7857	37.5	-0.31619	19107	46490	4198	652	1	21	34
lomita-serra-intersection-2019-01-30_0	17.097	25.665	17.379	28.018	72.478	5.5556	33.333	-3.7299	5770	14824	2191	58	2	12	22
meyer-green-2019-03-16_1	12.623	24.503	13.123	19.719	74.935	3.125	21.875	-2.2639	4616	18793	1544	117	2	14	48
nvidia-aud-2019-01-25_0	19.057	26.121	20.189	33.661	71.417	6.4516	43.548	-5.8117	11151	21976	4463	375	4	27	31
nvidia-aud-2019-04-18_1	35.083	19.254	35.534	49.762	77.765	25.926	29.63	-5.0982	3973	4011	1136	36	7	8	12
nvidia-aud-2019-04-18_2	23.885	27.531	24.514	52.976	65.051	23.529	44.118	-14.506	5394	4788	2898	64	8	15	11
quarry-road-2019-02-28_0	3.2668	25.639	3.8592	33.734	53.034	4.1667	41.667	-21.818	1993	3915	1765	35	1	10	13
serra-street-2019-01-30_0	8.7754	26.063	9.2171	17	68.596	2.6549	25.664	-3.7937	7313	35705	3348	190	3	29	81
stlc-111-2019-04-19_1	57.946	18.115	59.228	66.485	90.159	20.833	70.833	3.5051	4150	2092	453	80	5	17	2
stlc-111-2019-04-19_2	52.397	18.66	53.361	65.399	84.454	15.789	57.895	-0.79892	2374	1256	437	35	3	11	5
tressider-2019-03-16_2	16.68	24.437	17.122	23.043	79.56	7.2727	25.455	-0.73197	6142	20513	1578	118	4	14	37
tressider-2019-04-26_0	15.104	25.025	16.267	20.967	81.689	1.1628	33.721	-0.61632	25103	94622	5627	1393	3	87	168
tressider-2019-04-26_1	26.618	23.697	27.215	28.919	94.435	5.9361	36.073	4.5518	55681	136857	3281	1150	13	79	127
tressider-2019-04-26_3	19.925	26.065	21.302	27.779	81.093	2.5641

Leaderboard

Instructions

2D Tracking Submissions