Stable at Any Speed: Speed-Driven Multi-Object Tracking with Learnable Kalman Filtering

TL;DR

This paper introduces a Speed-Guided Learnable Kalman Filter for multi-object tracking that dynamically adapts to ego-vehicle speed, significantly enhancing tracking stability and accuracy in high-speed scenarios.

Contribution

It proposes a novel SG-LKF framework with MotionScaleNet to adapt uncertainty modeling based on ego-vehicle speed, improving multi-object tracking performance.

Findings

Achieves 79.59% HOTA on KITTI 2D MOT

Attains 82.03% HOTA on KITTI 3D MOT

Outperforms existing methods on nuScenes 3D MOT

Abstract

Multi-object tracking (MOT) enables autonomous vehicles to continuously perceive dynamic objects, supplying essential temporal cues for prediction, behavior understanding, and safe planning. However, conventional tracking-by-detection methods typically rely on static coordinate transformations based on ego-vehicle poses, disregarding ego-vehicle speed-induced variations in observation noise and reference frame changes, which degrades tracking stability and accuracy in dynamic, high-speed scenarios. In this paper, we investigate the critical role of ego-vehicle speed in MOT and propose a Speed-Guided Learnable Kalman Filter (SG-LKF) that dynamically adapts uncertainty modeling to ego-vehicle speed, significantly improving stability and accuracy in highly dynamic scenarios. Central to SG-LKF is MotionScaleNet (MSNet), a decoupled token-mixing and channel-mixing MLP that adaptively predicts key parameters of SG-LKF. To enhance inter-frame association and trajectory continuity, we introduce a self-supervised trajectory consistency loss jointly optimized with semantic and positional constraints. Extensive experiments show that SG-LKF ranks first among all vision-based methods on KITTI 2D MOT with 79.59% HOTA, delivers strong results on KITTI 3D MOT with 82.03% HOTA, and outperforms SimpleTrack by 2.2% AMOTA on nuScenes 3D MOT.