Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Efficient Visual Tracking with Exemplar Transformers

About

The design of more complex and powerful neural network models has significantly advanced the state-of-the-art in visual object tracking. These advances can be attributed to deeper networks, or the introduction of new building blocks, such as transformers. However, in the pursuit of increased tracking performance, runtime is often hindered. Furthermore, efficient tracking architectures have received surprisingly little attention. In this paper, we introduce the Exemplar Transformer, a transformer module utilizing a single instance level attention layer for realtime visual object tracking. E.T.Track, our visual tracker that incorporates Exemplar Transformer modules, runs at 47 FPS on a CPU. This is up to 8x faster than other transformer-based models. When compared to lightweight trackers that can operate in realtime on standard CPUs, E.T.Track consistently outperforms all other methods on the LaSOT, OTB-100, NFS, TrackingNet, and VOT-ST2020 datasets. Code and models are available at https://github.com/pblatter/ettrack.

Philippe Blatter, Menelaos Kanakis, Martin Danelljan, Luc Van Gool• 2021

Related benchmarks

TaskDatasetResultRank
Visual Object TrackingTrackingNet (test)
Normalized Precision (Pnorm)80.3
460
Object TrackingLaSoT
AUC59.1
333
Object TrackingTrackingNet
Precision (P)70.6
225
Visual Object TrackingUAV123 (test)
AUC62.3
188
Visual Object TrackingUAV123
AUC0.623
165
Visual Object TrackingOTB-100
AUC67.8
136
Visual Object TrackingNfS
AUC0.59
112
Object TrackingGOT-10k
AO56.6
74
Visual Object TrackingNFS (Need for Speed) 30 FPS (test)
AUC59
54
Visual Object TrackingVOT ST 2020
Robustness0.741
23
Showing 10 of 12 rows

Other info

Code

Follow for update