Box-Aware Feature Enhancement for Single Object Tracking on Point Clouds

TL;DR

This paper introduces a box-aware feature enhancement method for 3D single object tracking on point clouds, leveraging ground truth bounding boxes to improve accuracy and robustness against occlusion and sparsity.

Contribution

The paper proposes the BoxCloud representation and a box-aware feature fusion module, integrated into an existing tracker to significantly improve tracking performance.

Findings

Outperforms state-of-the-art on KITTI and NuScenes benchmarks

Achieves 15.2% improvement in precision

Runs approximately 20% faster

Abstract

Current 3D single object tracking approaches track the target based on a feature comparison between the target template and the search area. However, due to the common occlusion in LiDAR scans, it is non-trivial to conduct accurate feature comparisons on severe sparse and incomplete shapes. In this work, we exploit the ground truth bounding box given in the first frame as a strong cue to enhance the feature description of the target object, enabling a more accurate feature comparison in a simple yet effective way. In particular, we first propose the BoxCloud, an informative and robust representation, to depict an object using the point-to-box relation. We further design an efficient box-aware feature fusion module, which leverages the aforementioned BoxCloud for reliable feature matching and embedding. Integrating the proposed general components into an existing model P2B, we construct a superior box-aware tracker (BAT). Experiments confirm that our proposed BAT outperforms the previous state-of-the-art by a large margin on both KITTI and NuScenes benchmarks, achieving a 15.2% improvement in terms of precision while running ~20% faster.