🤖 AI Summary
To address low data association accuracy and weak temporal consistency in online multi-object tracking (MOT) for LiDAR point clouds, this paper proposes a modular tracker integrating graph optimization and self-attention. Methodologically, we design an end-to-end differentiable data association module that, for the first time in pure LiDAR online tracking, enables learning-based generation of structured matching score maps. We further introduce a graph neural network to model high-order relationships between tracks and detections, enhance feature interaction via self-attention, and integrate Kalman filtering to ensure temporal state consistency. Our key contribution is establishing a novel “learning-driven + graph-optimization” collaborative association paradigm. Experiments demonstrate that our method achieved top performance on the KITTI car online tracking benchmark and has consistently ranked second—outperforming all existing pure LiDAR and LiDAR-camera fusion approaches by significant margins.
📝 Abstract
Online multi-object tracking (MOT) plays a pivotal role in autonomous systems. The state-of-the-art approaches usually employ a tracking-by-detection method, and data association plays a critical role. This paper proposes a learning and graph-optimized (LEGO) modular tracker to improve data association performance in the existing literature. The proposed LEGO tracker integrates graph optimization and self-attention mechanisms, which efficiently formulate the association score map, facilitating the accurate and efficient matching of objects across time frames. To further enhance the state update process, the Kalman filter is added to ensure consistent tracking by incorporating temporal coherence in the object states. Our proposed method utilizing LiDAR alone has shown exceptional performance compared to other online tracking approaches, including LiDAR-based and LiDAR-camera fusion-based methods. LEGO ranked 1st at the time of submitting results to KITTI object tracking evaluation ranking board and remains 2nd at the time of submitting this paper, among all online trackers in the KITTI MOT benchmark for cars1
