A Graph Neural Network Approach for Solving the Ranked Assignment Problem in Multi-Object Tracking

arXiv cs.RO / 4/3/2026

💬 OpinionSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper targets the ranked assignment problem within Multi-Object Tracking (MOT), where truncation of hypotheses is typically handled by Murty’s algorithm or Gibbs sampling but with trade-offs in complexity or accuracy.
It proposes RAPNet, a Ranked Assignment Prediction Graph Neural Network that formulates the data association ranked assignment task as bipartite graphs for learning-based prediction.
The approach leverages Graph Neural Networks to improve how ranked assignments are computed under the δ-GLMB filtering framework.
Experiments report that RAPNet achieves higher accuracy than the Gibbs sampler when compared against Murty’s algorithm and Gibbs sampling.

Abstract

Associating measurements with tracks is a crucial step in Multi-Object Tracking (MOT) to guarantee the safety of autonomous vehicles. To manage the exponentially growing number of track hypotheses, truncation becomes necessary. In the

\delta

-Generalized Labeled Multi-Bernoulli (

\delta

-GLMB) filter application, this truncation typically involves the ranked assignment problem, solved by Murty's algorithm or the Gibbs sampling approach, both with limitations in terms of complexity or accuracy, respectively. With the motivation to improve these limitations, this paper addresses the ranked assignment problem arising from data association tasks with an approach that employs Graph Neural Networks (GNNs). The proposed Ranked Assignment Prediction Graph Neural Network (RAPNet) uses bipartite graphs to model the problem, harnessing the computational capabilities of deep learning. The conclusive evaluation compares the RAPNet with Murty's algorithm and the Gibbs sampler, showing accuracy improvements compared to the Gibbs sampler.