MOT Challenge

MOT17-08-FRCNN

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:

MOT17 |

Short name:

MO3TR_Plus

Detector:

Private

Description:

This is official implementation of MO3TR.
This paper has been accepted as a Regular Paper in an upcoming issue of the Transactions on Pattern Analysis and Machine Intelligence (Tpami).

Reference:

T. Zhu, M. Hiller, M. Ehsanpour, R. Ma, T. Drummond, I. Reid, H. Rezatofighi. Looking Beyond Two Frames: End-to-End Multi-Object Tracking Using Spatial and Temporal Transformers. In IEEE Transactions on Pattern Analysis and Machine Intelligence(TPAMI), 2022.

Last submitted:

October 09, 2022 (3 years ago)

Published:

October 09, 2022 at 14:14:53 CET

Submissions:

Project page / code:

https://github.com/alanzty/MO3TR

Open source:

Hardware:

Gtx2080ti

Runtime:

11.0 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	78.6	72.4	60.5	1,128 (47.9)	399 (16.9)	24,873	93,174	83.5	95.0	58.1	63.5	62.3	79.5	68.6	78.1	82.6	1.4	2,808 (0.0)	3,201 (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	64.0	57.2	51.0	12	5	437	1,859	71.2	91.3	48.7	53.5	52.3	80.8	59.1	75.8	83.5	1.0	26	31
MOT17-01-FRCNN	64.0	57.2	51.0	12	5	437	1,859	71.2	91.3	48.7	53.5	52.3	80.8	59.1	75.8	83.5	1.0	26	31
MOT17-01-SDP	64.0	57.2	51.0	12	5	437	1,859	71.2	91.3	48.7	53.5	52.3	80.8	59.1	75.8	83.5	1.0	26	31
MOT17-03-DPM	91.4	85.2	69.7	140	0	4,152	4,690	95.5	96.0	67.5	72.4	71.2	81.6	78.1	78.5	82.2	2.8	108	153
MOT17-03-FRCNN	91.4	85.2	69.7	140	0	4,152	4,690	95.5	96.0	67.5	72.4	71.2	81.6	78.1	78.5	82.2	2.8	108	153
MOT17-03-SDP	91.4	85.2	69.7	140	0	4,152	4,690	95.5	96.0	67.5	72.4	71.2	81.6	78.1	78.5	82.2	2.8	108	153
MOT17-06-DPM	58.4	56.5	47.6	89	60	1,015	3,699	68.6	88.8	44.5	51.3	51.1	72.8	57.2	74.0	83.0	0.9	192	198
MOT17-06-FRCNN	58.4	56.5	47.6	89	60	1,015	3,699	68.6	88.8	44.5	51.3	51.1	72.8	57.2	74.0	83.0	0.9	192	198
MOT17-06-SDP	58.4	56.5	47.6	89	60	1,015	3,699	68.6	88.8	44.5	51.3	51.1	72.8	57.2	74.0	83.0	0.9	192	198
MOT17-07-DPM	73.4	56.4	49.1	28	4	533	3,840	77.3	96.1	40.4	60.0	42.3	80.9	64.1	79.7	83.4	1.1	127	141
MOT17-07-FRCNN	73.4	56.4	49.1	28	4	533	3,840	77.3	96.1	40.4	60.0	42.3	80.9	64.1	79.7	83.4	1.1	127	141
MOT17-07-SDP	73.4	56.4	49.1	28	4	533	3,840	77.3	96.1	40.4	60.0	42.3	80.9	64.1	79.7	83.4	1.1	127	141
MOT17-08-DPM	62.9	45.1	42.7	33	5	1,254	6,292	70.2	92.2	35.1	53.6	37.9	75.0	58.4	76.7	82.8	2.0	286	314
MOT17-08-FRCNN	62.9	45.1	42.7	33	5	1,254	6,292	70.2	92.2	35.1	53.6	37.9	75.0	58.4	76.7	82.8	2.0	286	314
MOT17-08-SDP	62.9	45.1	42.7	33	5	1,254	6,292	70.2	92.2	35.1	53.6	37.9	75.0	58.4	76.7	82.8	2.0	286	314
MOT17-12-DPM	70.8	64.1	53.9	46	17	310	2,180	74.8	95.4	48.8	59.7	67.3	60.5	64.1	81.7	85.7	0.3	43	54
MOT17-12-FRCNN	70.8	64.1	53.9	46	17	310	2,180	74.8	95.4	48.8	59.7	67.3	60.5	64.1	81.7	85.7	0.3	43	54
MOT17-12-SDP	70.8	64.1	53.9	46	17	310	2,180	74.8	95.4	48.8	59.7	67.3	60.5	64.1	81.7	85.7	0.3	43	54
MOT17-14-DPM	50.0	54.6	41.4	28	42	590	8,498	54.0	94.4	41.7	41.3	44.4	77.7	43.8	76.6	82.1	0.8	154	176
MOT17-14-FRCNN	50.0	54.6	41.4	28	42	590	8,498	54.0	94.4	41.7	41.3	44.4	77.7	43.8	76.6	82.1	0.8	154	176
MOT17-14-SDP	50.0	54.6	41.4	28	42	590	8,498	54.0	94.4	41.7	41.3	44.4	77.7	43.8	76.6	82.1	0.8	154	176

Raw data:

download

FastTracker

MO3TR_Plus

AFN17

MO3TR_Plus: Looking Beyond Two Frames: End-to-End Multi-Object Tracking Using Spatial and Temporal Transformers