SearchTrack: Multiple Object Tracking with Object-Customized Search and Motion-Aware Features

TL;DR

SearchTrack introduces a novel multiple object tracking method that combines object-customized search with motion-aware features, leveraging Kalman filters and dynamic convolutions to improve association accuracy in MOT tasks.

Contribution

The paper proposes a new MOT approach that integrates object-specific search and motion-aware features, enhancing association performance over existing methods.

Findings

Outperforms competitive methods on MOTS and MOT datasets.

Achieves 71.5 HOTA on KITTI MOTS for cars.

Achieves 57.6 HOTA on KITTI MOTS for pedestrians.

Abstract

The paper presents a new method, SearchTrack, for multiple object tracking and segmentation (MOTS). To address the association problem between detected objects, SearchTrack proposes object-customized search and motion-aware features. By maintaining a Kalman filter for each object, we encode the predicted motion into the motion-aware feature, which includes both motion and appearance cues. For each object, a customized fully convolutional search engine is created by SearchTrack by learning a set of weights for dynamic convolutions specific to the object. Experiments demonstrate that our SearchTrack method outperforms competitive methods on both MOTS and MOT tasks, particularly in terms of association accuracy. Our method achieves 71.5 HOTA (car) and 57.6 HOTA (pedestrian) on the KITTI MOTS and 53.4 HOTA on MOT17. In terms of association accuracy, our method achieves state-of-the-art performance among 2D online methods on the KITTI MOTS. Our code is available at https://github.com/qa276390/SearchTrack.