PermaTrackPr: Learning to Track with Object Permanence


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.

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.

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.

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

Benchmark:

MOT17 |

Short name:

PermaTrackPr

Detector:

Private

Description:

This is a variant of our model evaluated with private detections.

Reference:

P. Tokmakov, J. Li, W. Burgard, A. Gaidon. Learning to Track with Object Permanence. In ICCV, 2021.

Last submitted:

June 12, 2021 (3 years ago)

Published:

June 16, 2021 at 01:27:37 CET

Submissions:

1

Project page / code:

n/a

Open source:

No

Hardware:

Tesla V100

Runtime:

11.9 Hz

Benchmark performance:

Sequence MOTA IDF1 HOTA MT ML FP FN Rcll Prcn AssA DetA AssRe AssPr DetRe DetPr LocA FAF ID Sw. Frag
MOT1773.868.955.51,032 (43.8)405 (17.2)28,998115,10479.693.953.158.559.870.163.975.481.41.63,699 (0.0)6,132 (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-DPM50.948.239.21388312,30364.383.335.943.043.259.750.164.979.91.83262
MOT17-01-FRCNN50.948.239.21388312,30364.383.335.943.043.259.750.164.979.91.83262
MOT17-01-SDP50.948.239.21388312,30364.383.335.943.043.259.750.164.979.91.83262
MOT17-03-DPM89.582.764.913312,2998,58791.897.761.668.967.475.873.378.081.21.5141401
MOT17-03-FRCNN89.582.764.913312,2998,58791.897.761.668.967.475.873.378.081.21.5141401
MOT17-03-SDP89.582.764.913312,2998,58791.897.761.668.967.475.873.378.081.21.5141401
MOT17-06-DPM61.148.039.482447073,65569.092.031.350.349.844.855.173.581.00.6220302
MOT17-06-FRCNN61.148.039.482447073,65569.092.031.350.349.844.855.173.581.00.6220302
MOT17-06-SDP61.148.039.482447073,65569.092.031.350.349.844.855.173.581.00.6220302
MOT17-07-DPM59.847.340.42642,1924,36374.285.133.550.239.059.258.667.280.14.4240385
MOT17-07-FRCNN59.847.340.42642,1924,36374.285.133.550.239.059.258.667.280.14.4240385
MOT17-07-SDP59.847.340.42642,1924,36374.285.133.550.239.059.258.667.280.14.4240385
MOT17-08-DPM51.246.139.824158569,25556.293.337.143.643.860.846.777.583.41.4188268
MOT17-08-FRCNN51.246.139.824158569,25556.293.337.143.643.860.846.777.583.41.4188268
MOT17-08-SDP51.246.139.824158569,25556.293.337.143.643.860.846.777.583.41.4188268
MOT17-12-DPM46.757.249.327281,5173,06564.678.753.345.863.170.855.367.385.51.74177
MOT17-12-FRCNN46.757.249.327281,5173,06564.678.753.345.863.170.855.367.385.51.74177
MOT17-12-SDP46.757.249.327281,5173,06564.678.753.345.863.170.855.367.385.51.74177
MOT17-14-DPM52.552.540.939351,2647,14061.490.037.744.743.061.649.272.281.21.7371549
MOT17-14-FRCNN52.552.540.939351,2647,14061.490.037.744.743.061.649.272.281.21.7371549
MOT17-14-SDP52.552.540.939351,2647,14061.490.037.744.743.061.649.272.281.21.7371549

Raw data:


PermaTrackPr