ADUULM-TTB: A Scalable, Generic, and Efficient Multi-Sensor Multi-Object Tracking Toolbox

Tracking multiple objects in dynamic environments is a crucial task in various applications, including automated driving, military surveillance, and robotics. While significant efforts have been made to develop open-source libraries, such as the Stone-Soup framework, these primarily focus on offline evaluation and benchmarking. However, a gap remains in the availability of a high-performance, scalable, and generic tracking solution suitable for real-world deployment. In this work, we present a multi-sensor multi-object tracking toolbox, referred to as ADUULM-TTB, designed for real-world applications. Our framework combines scalability and generality with an efficient, high-performance implementation, making it suitable for real-time usage. To enhance usability, we integrate a direct interface with the Stone-Soup framework, allowing for seamless evaluation and integration. The ADUULM-TTB is open-source and available at https://github.com/uulm-mrm/aduulm_ttb.