Classification-Aided Robust Multiple Target Tracking Using Neural Enhanced Message Passing

We address the challenge of tracking an unknown number of targets in strong clutter environments using measurements provided by a radar sensor. Leveraging the range-Doppler spectra information, we identify the measurement classes, which serve as additional information to enhance clutter rejection and data association, thus bolstering the robustness of target tracking. We first introduce a novel neural enhanced message passing approach, where the belief obtained by the unified message passing is fed into the neural network as additional information. The output belief is then utilized to refine the original belief. Then, we propose a classification-aided robust multitarget tracking algorithm employing the neural enhanced message passing technique. This algorithm is comprised of three modules: a factor graph module, a neural network module, and a Dempster-Shafer combination module. The factor graph module uses the factor graph to represent the statistical model of the problem and infers the target kinematic state, visibility state, and data association based on spatial measurements. The neural network module is employed to extract feature of range-Doppler spectra and derive belief on whether a measurement is target-generated or clutter-generated. The Dempster-Shafer module is used to fuse the beliefs obtained from both the factor graph and the neural network. As a result, our proposed algorithm adopts a model-and-data-driven framework, effectively enhancing clutter suppression and data association, leading to significant improvements in multiple target tracking performance. We validate the effectiveness of our approach using both simulated and real data scenarios, demonstrating its capability to handle challenging tracking scenarios in practical radar applications.