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

MOT17-14-FRCNN

MOT17-01-SDP

MOT17-03-SDP

MOT17-06-SDP

MOT17-07-SDP

MOT17-08-SDP

MOT17-12-SDP

MOT17-14-SDP

Benchmark:

MOT20 | MOT17 |

Short name:

Swin_JDE

Detector:

Private

Description:

Swin-JDE is a novel anchor-free joint detection and embedding (JDE) MOT method based on Transformer architecture, which provides excellent tracking performance in crowded scenes.

Reference:

C.-Y. Tsai, G.-Y. Shen and H. Nisar, "Swin-JDE: Joint Detection and Embedding Multi-Object Tracking Based on Swin-Transformer," Engineering Applications of Artificial Intelligence, Vol. 119, No. 105770, pp. 1-16, 2023.

Last submitted:

June 29, 2022 (2 years ago)

Published:

August 03, 2022 at 04:55:58 CET

Submissions:

Project page / code:

n/a

Open source:

Hardware:

3.6GHZ, 4 Cores

Runtime:

4.5 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	72.3	70.7	57.8	888 (37.7)	537 (22.8)	20,268	133,440	76.4	95.5	57.4	58.5	61.8	78.8	62.8	78.5	83.0	1.1	2,679 (0.0)	3,903 (0.0)

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	57.7	60.6	49.5	8	8	78	2,636	59.1	98.0	51.9	47.4	56.4	81.5	49.3	81.8	83.9	0.2	15	40
MOT17-01-FRCNN	57.7	60.6	49.5	8	8	78	2,636	59.1	98.0	51.9	47.4	56.4	81.5	49.3	81.8	83.9	0.2	15	40
MOT17-01-SDP	57.7	60.6	49.5	8	8	78	2,636	59.1	98.0	51.9	47.4	56.4	81.5	49.3	81.8	83.9	0.2	15	40
MOT17-03-DPM	89.5	81.0	67.0	140	0	5,343	5,456	94.8	94.9	63.1	71.4	67.9	79.3	78.1	78.2	83.1	3.6	159	344
MOT17-03-FRCNN	89.5	81.0	67.0	140	0	5,343	5,456	94.8	94.9	63.1	71.4	67.9	79.3	78.1	78.2	83.1	3.6	159	344
MOT17-03-SDP	89.5	81.0	67.0	140	0	5,343	5,456	94.8	94.9	63.1	71.4	67.9	79.3	78.1	78.2	83.1	3.6	159	344
MOT17-06-DPM	59.0	60.3	47.9	69	64	313	4,314	63.4	96.0	47.7	48.4	53.3	76.4	51.3	77.6	81.9	0.3	210	231
MOT17-06-FRCNN	59.0	60.3	47.9	69	64	313	4,314	63.4	96.0	47.7	48.4	53.3	76.4	51.3	77.6	81.9	0.3	210	231
MOT17-06-SDP	59.0	60.3	47.9	69	64	313	4,314	63.4	96.0	47.7	48.4	53.3	76.4	51.3	77.6	81.9	0.3	210	231
MOT17-07-DPM	61.5	53.7	43.7	15	9	209	6,201	63.3	98.1	39.2	49.0	41.7	75.2	51.4	79.6	82.5	0.4	98	172
MOT17-07-FRCNN	61.5	53.7	43.7	15	9	209	6,201	63.3	98.1	39.2	49.0	41.7	75.2	51.4	79.6	82.5	0.4	98	172
MOT17-07-SDP	61.5	53.7	43.7	15	9	209	6,201	63.3	98.1	39.2	49.0	41.7	75.2	51.4	79.6	82.5	0.4	98	172
MOT17-08-DPM	38.9	44.4	37.7	16	26	292	12,536	40.7	96.7	43.4	32.9	46.0	80.7	34.1	81.1	84.2	0.5	88	166
MOT17-08-FRCNN	38.9	44.4	37.7	16	26	292	12,536	40.7	96.7	43.4	32.9	46.0	80.7	34.1	81.1	84.2	0.5	88	166
MOT17-08-SDP	38.9	44.4	37.7	16	26	292	12,536	40.7	96.7	43.4	32.9	46.0	80.7	34.1	81.1	84.2	0.5	88	166
MOT17-12-DPM	60.9	65.9	52.5	30	22	357	2,991	65.5	94.1	54.4	50.8	59.0	78.1	54.3	78.1	83.3	0.4	45	76
MOT17-12-FRCNN	60.9	65.9	52.5	30	22	357	2,991	65.5	94.1	54.4	50.8	59.0	78.1	54.3	78.1	83.3	0.4	45	76
MOT17-12-SDP	60.9	65.9	52.5	30	22	357	2,991	65.5	94.1	54.4	50.8	59.0	78.1	54.3	78.1	83.3	0.4	45	76
MOT17-14-DPM	41.6	51.6	35.9	18	50	164	10,346	44.0	98.0	38.4	33.7	40.5	75.4	34.9	77.7	80.9	0.2	278	272
MOT17-14-FRCNN	41.6	51.6	35.9	18	50	164	10,346	44.0	98.0	38.4	33.7	40.5	75.4	34.9	77.7	80.9	0.2	278	272
MOT17-14-SDP	41.6	51.6	35.9	18	50	164	10,346	44.0	98.0	38.4	33.7	40.5	75.4	34.9	77.7	80.9	0.2	278	272

Raw data:

download

SUT

Swin_JDE

AFN17

Swin_JDE: Joint Detection and Embedding Multi-Object Tracking Based on Swin-Transformer