MHT with Neural Gating using Bilinear LSTM

MOT16-01 MOT16-03 MOT16-06 MOT16-07 MOT16-08 MOT16-12 MOT16-14

Short name:

MHT_bLSTM6

Benchmark:

Description:

n/a

Hardware:

Titan X GPU, 3.6GHZ CPU

Detector:

Public

Processing:

Batch

Last submitted:

July 26, 2018 (3 months ago)

Published:

July 26, 2018 at 09:04:30 CET

Submissions:

3

Open source:

No

Project page / code:

n/a

Reference:

C. Kim, F. Li, J. Rehg. Multi-object Tracking with Neural Gating Using Bilinear LSTM. In ECCV, 2018.

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
42.175.92.014.9 % 44.4 % 11,63793,1727531,156Titan X GPU, 3.6GHZ CPUPublic
IDF1ID PrecisionID Recall
47.867.237.1

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
MOT16-0127.536.571.61.72326.1 % 39.1 % 7853,8133963
MOT16-0349.453.475.84.914821.6 % 20.3 % 7,29945,264300532
MOT16-0642.245.774.60.922117.6 % 48.0 % 1,0675,50791132
MOT16-0734.941.175.11.0547.4 % 44.4 % 49710,022113158
MOT16-0833.036.779.51.16317.5 % 39.7 % 67410,44787111
MOT16-1238.849.777.60.88615.1 % 46.5 % 7064,3403143
MOT16-1421.731.075.40.81644.9 % 62.8 % 60913,77992117

Raw data:


MHT_bLSTM6