MOT Challenge

Video not available.

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:

CVPR 2019 Detection Challenge |

Short name:

ViPeD_19

Description:

n/a

Reference:

G. Amato, L. Ciampi, F. Falchi, C. Gennaro, N. Messina. Learning pedestrian detection from virtual worlds. In International Conference on Image Analysis and Processing, 2019.

Last submitted:

June 13, 2019 (4 years ago)

Published:

July 05, 2019 at 23:52:29 CET

Submissions:

Project page / code:

n/a

Open source:

Yes

Hardware:

RTX 2080Ti

Runtime:

11.2 Hz

Benchmark performance:

Sequence	AP	MODA	MODP	FAF	TP	FP	FN	Rcll	Prcn	F1
CVPR 2019 Detection Challenge	0.73	39.0	71.5	35.7	308,545	159,983	71,997	81.1	65.9	72.7

Detailed performance:

Sequence	AP	MODA	MODP	FAF	TP	FP	FN	Rcll	Prcn	F1
CVPR19-04	0.77	57.4	71.0	32.3	221,665	67,191	47,282	82.4	76.7	79.5
CVPR19-06	0.65	-7.8	72.4	51.0	46,539	51,409	16,186	74.2	47.5	57.9
CVPR19-07	0.89	65.9	74.9	7.0	14,858	4,109	1,460	91.1	78.3	84.2
CVPR19-08	0.61	-36.2	71.7	46.2	25,483	37,274	7,069	78.3	40.6	53.5

Raw data:

download

ViPeD_19

ViPeD_19

MOT19_2

ViPeD_19: YOLO detector trained using JTA