Speaker
Description
This abstract outlines my current research for my PhD thesis, focusing specifically on creating a synthetic dataset for multi-camera multi-object tracking (MCMOT) within logistics applications.
Motivation: Tracking moving assets such as trucks, trailers, or containers in logistics yards is crucial for developing digital twins, measuring key performance indicators, and enhancing operational efficiency. Effective MCMOT is essential yet challenging due to the presence of visually similar objects, frequent long-term occlusions, and extensive periods of low activity or inactivity.
Research Gap: Currently, there is a significant lack of open, logistics-specific datasets tailored for MCMOT applications. Most available datasets predominantly address urban or surveillance scenarios and rarely cover large-scale environments with overlapping fields of view for warehouses or cross-docking facilities. This absence severely restricts benchmarking, evaluation, and development of robust tracking algorithms for logistics scenarios, that don't rely heavily on re-identification.
Method: To address this gap, my research will involve the systematic creation of a large-scale synthetic dataset specifically designed for logistics yard scenarios. Characteristics of real-world warehouse and cross-dock environments will be realistically designed and simulated using Blender, building upon prior research on automated generation of labeled MCMOT datasets (https://ieeexplore.ieee.org/document/10710720). This dataset will focus explicitly on the tracking challenge, providing precise, automatically generated annotations to facilitate targeted experimentation and algorithmic development. Detection and segmentation challenges, e.g. addressing the sim-2-real gap, will be considered secondary to ensure primary focus remains on multi-camera tracking performance.
This targeted dataset creation effort aims to advance open research capabilities in logistics-specific MCMOT, enabling future development of more robust, efficient, and resource-aware tracking solutions.
