AGT: Adapted Graph Tracking

MOT16-01


Benchmark:

MOT17 | MOT16 |

Short name:

AGT

Detector:

Public

Description:

graph tracking with self adapted weights.

Reference:

Alibaba DAMO Academy city-brain team

Last submitted:

August 17, 2020 (2 months ago)

Published:

July 24, 2020 at 12:51:25 CET

Submissions:

4

Project page / code:

n/a

Open source:

No

Hardware:

Tesla-V100 32 G, 2 Core

Runtime:

6.6 Hz

Benchmark performance:

Sequence MOTA IDF1 MOTP MT ML FP FN Recall Precision FAF ID Sw. Frag
MOT1674.867.281.2342 (45.1)112 (14.8)12,17332,46282.292.52.11,239 (15.1)2,647 (32.2)

Detailed performance:

Sequence MOTA IDF1 MOTP MT ML FP FN Recall Precision FAF ID Sw. Frag
MOT16-0158.149.181.81043902,25164.891.40.939107
MOT16-0389.277.482.413104,1506,94793.495.92.8243670
MOT16-0663.361.280.188436463,40470.592.60.5189270
MOT16-0764.051.880.12221,2574,47672.690.42.5143385
MOT16-0846.140.779.62123,3145,38267.877.45.3321469
MOT16-1258.664.980.127166132,77666.590.00.746176
MOT16-1449.852.674.743451,8037,22660.986.22.4258570

Raw data: