Exploring Simple 3D Multi-Object Tracking for Autonomous Driving

TL;DR

This paper introduces SimTrack, an end-to-end trainable model for 3D multi-object tracking in LiDAR data that simplifies existing methods by removing heuristic matching, achieving competitive results on large datasets.

Contribution

Proposes SimTrack, a unified model for joint detection and tracking in 3D LiDAR data, eliminating heuristic matching steps and improving simplicity and efficiency.

Findings

Achieves competitive performance on nuScenes and Waymo datasets.

Successfully removes heuristic matching in the tracking pipeline.

Demonstrates the effectiveness of end-to-end training for 3D multi-object tracking.

Abstract

3D multi-object tracking in LiDAR point clouds is a key ingredient for self-driving vehicles. Existing methods are predominantly based on the tracking-by-detection pipeline and inevitably require a heuristic matching step for the detection association. In this paper, we present SimTrack to simplify the hand-crafted tracking paradigm by proposing an end-to-end trainable model for joint detection and tracking from raw point clouds. Our key design is to predict the first-appear location of each object in a given snippet to get the tracking identity and then update the location based on motion estimation. In the inference, the heuristic matching step can be completely waived by a simple read-off operation. SimTrack integrates the tracked object association, newborn object detection, and dead track killing in a single unified model. We conduct extensive evaluations on two large-scale datasets: nuScenes and Waymo Open Dataset. Experimental results reveal that our simple approach compares favorably with the state-of-the-art methods while ruling out the heuristic matching rules.