Context-Aware Sensor Modeling for Asynchronous Multi-Sensor Tracking in Stone Soup

TL;DR

This paper introduces DetectorContext, an extension to the Stone Soup framework, enabling context-aware modeling of detection probability and clutter, which improves multi-sensor tracking performance in asynchronous, heterogeneous sensor environments.

Contribution

It presents DetectorContext, a novel abstraction that incorporates state-dependent detection and clutter modeling into existing probabilistic trackers without altering their core algorithms.

Findings

Restores stable sensor fusion in asynchronous environments

Significantly improves HOTA and GOSPA metrics

Does not increase false track counts

Abstract

Multi-sensor tracking in the real world involves asynchronous sensors with partial coverage and heterogeneous detection performance. Although probabilistic tracking methods permit detection probability and clutter intensity to depend on state and sensing context, many practical frameworks enforce globally uniform observability assumptions. Under multi-rate and partially overlapping sensing, this simplification causes repeated non-detections from high-rate sensors to erode tracks visible only to low-rate sensors, potentially degrading fusion performance. We introduce DetectorContext, an abstraction for the open-source multi-target tracking framework Stone Soup. DetectorContext exposes detection probability and clutter intensity as state-dependent functions evaluated during hypothesis formation. The abstraction integrates with existing probabilistic trackers without modifying their update equations. Experiments on asynchronous radar-lidar data demonstrate that context-aware modeling restores stable fusion and significantly improves HOTA and GOSPA performance without increasing false tracks.