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

2026-04-02

AI Summary

This work addresses the combinatorial challenge of ranked assignment in multi-object tracking, which suffers from exponential growth in hypothesis space and remains difficult to solve efficiently. Conventional approaches such as Murty’s algorithm and Gibbs sampling are limited by either high computational complexity or insufficient accuracy. To overcome these limitations, this paper introduces graph neural networks (GNNs) to this problem for the first time, proposing a novel deep learning architecture named RAPNet that formulates the assignment task as a bipartite graph matching problem. RAPNet achieves a favorable trade-off between assignment accuracy and computational efficiency. Experimental results demonstrate that RAPNet significantly outperforms Gibbs sampling in accuracy while exhibiting lower computational complexity than Murty’s algorithm, thereby effectively breaking through the performance bottlenecks of traditional methods.

Technology Category

• Computer Vision: Motion & Tracking
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Machine Learning: Graph-based Machine Learning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Topic discovery and tracking

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 $δ$-Generalized Labeled Multi-Bernoulli ($δ$-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.

Problem

• Ranked Assignment Problem
• Multi-Object Tracking
• Data Association
• δ-GLMB filter

Innovation

• Graph Neural Network
• Ranked Assignment Problem
• Multi-Object Tracking
• RAPNet
• Data Association