AAA: Adaptive Aggregation of Arbitrary Online Trackers with Theoretical Performance Guarantee

19 Sep 2020  ·  Heon Song, Daiki Suehiro, Seiichi Uchida ·

For visual object tracking, it is difficult to realize an almighty online tracker due to the huge variations of target appearance depending on an image sequence. This paper proposes an online tracking method that adaptively aggregates arbitrary multiple online trackers. The performance of the proposed method is theoretically guaranteed to be comparable to that of the best tracker for any image sequence, although the best expert is unknown during tracking. The experimental study on the large variations of benchmark datasets and aggregated trackers demonstrates that the proposed method can achieve state-of-the-art performance. The code is available at https://github.com/songheony/AAA-journal.

PDF Abstract

Datasets


Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Benchmark
Visual Object Tracking OTB-2015 AAA Precision 0.91 # 2
AUC 0.70 # 6
Visual Object Tracking TempleColor128 AAA AUC 0.62 # 1
Precision 0.84 # 1

Methods


No methods listed for this paper. Add relevant methods here