Efficient Visual Tracking with Exemplar Transformers

TL;DR

This paper introduces the Exemplar Transformer for real-time visual object tracking, achieving high speed and superior accuracy compared to existing transformer-based and lightweight trackers.

Contribution

The paper proposes the Exemplar Transformer module and E.T.Track, a fast and efficient visual tracker that outperforms existing methods in speed and accuracy.

Findings

E.T.Track runs at 47 FPS on a CPU, up to 8x faster than other transformer models.

E.T.Track outperforms all tested lightweight trackers on multiple datasets.

The Exemplar Transformer enables real-time tracking with high accuracy.

Abstract

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.