Improving Multi-Frame Data Association with Sparse Representations for Robust Near-Online Multi-Object Tracking

TUD-Crossing PETS09-S2L2 ETH-Jelmoli ETH-Linthescher ETH-Crossing AVG-TownCentre ADL-Rundle-1 ADL-Rundle-3 KITTI-16 KITTI-19

Short name:

LINF1

Benchmark:

Description:

n/a

Hardware:

2.7 GHZ, 8 Core

Detector:

Public

Processing:

Batch

Last submitted:

March 05, 2016 (1 year ago)

Published:

March 16, 2016 at 15:53:31 CET

Submissions:

1

Open source:

No

Project page / code:

n/a

Reference:

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
24.571.31.05.5 % 64.6 % 5,86440,2072987442.7 GHZ, 8 CorePublic
IDF1ID PrecisionID Recall
34.856.925.1

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
TUD-Crossing64.261.773.20.11338.5 % 15.4 % 12375822
PETS09-S2L237.734.170.20.9422.4 % 26.2 % 3935,497114202
ETH-Jelmoli36.552.673.50.44515.6 % 48.9 % 1691,4301135
ETH-Linthescher9.215.574.20.01971.5 % 87.8 % 408,067310
ETH-Crossing21.430.275.00.0267.7 % 76.9 % 977545
AVG-TownCentre11.120.770.40.22262.2 % 81.0 % 1066,241534
ADL-Rundle-114.440.371.56.13225.0 % 28.1 % 3,0564,87335106
ADL-Rundle-335.837.272.61.5449.1 % 25.0 % 9495,5225988
KITTI-1639.147.773.30.3170.0 % 23.5 % 589661226
KITTI-1927.842.466.00.6628.1 % 37.1 % 5923,23632186
Venice-118.532.172.81.1170.0 % 47.1 % 4803,2251530

Raw data:


LINF1
mLK