MOT Challenge

MOT17-01-DPM

MOT17-03-DPM

MOT17-06-DPM

MOT17-07-DPM

MOT17-08-DPM

MOT17-12-DPM

MOT17-14-DPM

MOT17-01-FRCNN

MOT17-03-FRCNN

MOT17-06-FRCNN

MOT17-07-FRCNN

MOT17-08-FRCNN

MOT17-12-FRCNN

Rendering of new sequences is currently deactivated due to heavy load.

Rendering of new sequences is currently deactivated due to heavy load.

Rendering of new sequences is currently deactivated due to heavy load.

Rendering of new sequences is currently deactivated due to heavy load.

Rendering of new sequences is currently deactivated due to heavy load.

Rendering of new sequences is currently deactivated due to heavy load.

Rendering of new sequences is currently deactivated due to heavy load.

Rendering of new sequences is currently deactivated due to heavy load.

Benchmark:

MOT20 | MOT17 | MOT16 |

Short name:

LPC_MOT

Detector:

Public

Description:

This approach is devised by Tsinghua University and AIBEE Inc.

Reference:

P. Dai, R. Weng, W. Choi, C. Zhang, Z. He, W. Ding. Learning a Proposal Classifier for Multiple Object tracking. In CVPR (Accepted), 2021.

Last submitted:

September 07, 2020 (4 years ago)

Published:

September 25, 2020 at 09:23:34 CET

Submissions:

Project page / code:

https://github.com/daip13/LPC_MOT

Open source:

Yes

Hardware:

GeForce GTX 1080 Ti

Runtime:

4.8 Hz

Benchmark performance:

Sequence	MOTA	IDF1	HOTA	MT	ML	FP	FN	Rcll	Prcn	AssA	DetA	AssRe	AssPr	DetRe	DetPr	LocA	FAF	ID Sw.	Frag
MOT17	59.0	66.8	51.5	703 (29.9)	798 (33.9)	23,102	206,948	63.3	93.9	56.0	47.7	62.7	74.0	51.0	75.7	80.9	1.3	1,122 (17.7)	1,943 (30.7)

Detailed performance:

Sequence	MOTA	IDF1	HOTA	MT	ML	FP	FN	Rcll	Prcn	AssA	DetA	AssRe	AssPr	DetRe	DetPr	LocA	FAF	ID Sw.	Frag
MOT17-01-DPM	47.3	55.2	45.4	8	11	104	3,285	49.1	96.8	54.6	37.7	61.5	73.8	39.1	77.2	80.5	0.2	12	15
MOT17-01-FRCNN	46.7	54.3	44.0	10	10	428	2,995	53.6	89.0	50.0	38.8	56.0	72.7	42.3	70.4	79.6	1.0	18	23
MOT17-01-SDP	49.9	55.0	45.2	9	9	186	3,025	53.1	94.8	51.6	39.6	55.8	77.3	41.8	74.7	80.0	0.4	20	28
MOT17-03-DPM	69.9	75.2	57.1	72	18	3,680	27,771	73.5	95.4	59.1	55.4	65.3	74.9	59.1	76.8	81.1	2.5	66	121
MOT17-03-FRCNN	71.1	76.9	57.7	72	17	3,497	26,650	74.5	95.7	59.8	55.8	65.4	75.6	59.6	76.5	80.8	2.3	68	151
MOT17-03-SDP	75.1	78.6	59.9	91	14	5,604	20,316	80.6	93.8	60.6	59.4	67.7	73.1	64.4	74.9	80.2	3.7	95	239
MOT17-06-DPM	55.7	58.7	47.0	75	86	892	4,290	63.6	89.4	46.7	47.4	66.9	59.6	52.1	73.2	81.4	0.7	44	85
MOT17-06-FRCNN	58.4	62.2	49.4	86	62	1,176	3,666	68.9	87.3	48.8	50.2	67.6	60.9	56.3	71.4	81.1	1.0	55	113
MOT17-06-SDP	57.9	61.8	48.9	92	63	1,245	3,656	69.0	86.7	48.1	50.0	67.6	60.5	56.4	70.9	81.1	1.0	62	103
MOT17-07-DPM	45.1	51.2	39.7	8	18	459	8,752	48.2	94.7	42.7	37.1	45.3	77.0	38.9	76.4	81.0	0.9	68	123
MOT17-07-FRCNN	44.8	49.7	38.8	8	19	517	8,733	48.3	94.0	41.0	37.0	43.5	77.3	39.0	75.9	81.0	1.0	76	116
MOT17-07-SDP	47.8	52.1	40.1	9	15	653	8,097	52.1	93.1	41.3	39.3	44.6	75.0	41.8	74.7	80.6	1.3	69	129
MOT17-08-DPM	33.8	45.4	40.5	18	37	516	13,432	36.4	93.7	56.0	29.4	61.5	76.6	30.6	79.0	83.7	0.8	29	45
MOT17-08-FRCNN	32.7	43.8	39.6	18	36	747	13,434	36.4	91.1	53.7	29.3	62.0	72.9	30.7	77.0	83.2	1.2	36	58
MOT17-08-SDP	34.2	46.6	40.8	20	37	841	13,032	38.3	90.6	54.6	30.5	61.7	73.6	32.2	76.3	83.1	1.3	36	58
MOT17-12-DPM	47.1	58.7	47.1	20	43	117	4,450	48.7	97.3	56.4	39.4	60.5	81.8	40.9	81.9	84.4	0.1	22	23
MOT17-12-FRCNN	46.1	57.9	46.3	18	43	81	4,572	47.2	98.1	55.8	38.5	60.0	79.7	39.8	82.5	84.4	0.1	16	18
MOT17-12-SDP	47.4	59.9	47.6	20	42	198	4,341	49.9	95.6	56.9	39.9	61.2	81.4	41.9	80.2	84.0	0.2	18	24
MOT17-14-DPM	35.1	45.7	33.3	13	78	427	11,473	37.9	94.3	38.8	28.7	42.1	71.5	29.9	74.4	79.9	0.6	99	127
MOT17-14-FRCNN	36.5	47.2	34.5	16	71	816	10,814	41.5	90.4	39.2	30.7	42.1	72.7	32.6	71.0	79.1	1.1	107	175
MOT17-14-SDP	39.5	49.1	36.0	20	69	918	10,164	45.0	90.1	39.4	33.1	42.7	71.2	35.4	70.9	79.2	1.2	106	169

Raw data:

download

MOT_FCG_plus

LPC_MOT

AFN17

LPC_MOT: Learning a Proposal Classifier for Multiple Object Tracking