End-to-end Flow Correlation Tracking with Spatial-temporal Attention

TL;DR

This paper introduces a deep learning-based tracking method that leverages optical flow and a novel spatial-temporal attention mechanism to enhance feature representation and tracking accuracy during challenging scenarios.

Contribution

It is the first to jointly train flow estimation and tracking in a unified deep learning framework, improving robustness against occlusion and deformation.

Findings

Achieves superior results on multiple challenging benchmarks.

Effectively utilizes flow information for improved tracking.

Introduces a novel spatial-temporal attention mechanism.

Abstract

Discriminative correlation filters (DCF) with deep convolutional features have achieved favorable performance in recent tracking benchmarks. However, most of existing DCF trackers only consider appearance features of current frame, and hardly benefit from motion and inter-frame information. The lack of temporal information degrades the tracking performance during challenges such as partial occlusion and deformation. In this work, we focus on making use of the rich flow information in consecutive frames to improve the feature representation and the tracking accuracy. Firstly, individual components, including optical flow estimation, feature extraction, aggregation and correlation filter tracking are formulated as special layers in network. To the best of our knowledge, this is the first work to jointly train flow and tracking task in a deep learning framework. Then the historical feature maps at predefined intervals are warped and aggregated with current ones by the guiding of flow. For adaptive aggregation, we propose a novel spatial-temporal attention mechanism. Extensive experiments are performed on four challenging tracking datasets: OTB2013, OTB2015, VOT2015 and VOT2016, and the proposed method achieves superior results on these benchmarks.