Sparse Message Passing Network with Feature Integration for Online Multiple Object Tracking

TL;DR

This paper introduces a simple online message passing network with feature integration and hierarchical sampling for online multiple object tracking, improving robustness and reducing identity switches compared to existing complex methods.

Contribution

The paper proposes a novel IoU-guided function for better long-term tracking and a hierarchical sampling strategy for sparser graph construction, enhancing tracking performance.

Findings

Outperforms many state-of-the-art methods in online MOT.

Improves long-term tracking and reduces identity switches.

Generalizes well to private detection based methods.

Abstract

Existing Multiple Object Tracking (MOT) methods design complex architectures for better tracking performance. However, without a proper organization of input information, they still fail to perform tracking robustly and suffer from frequent identity switches. In this paper, we propose two novel methods together with a simple online Message Passing Network (MPN) to address these limitations. First, we explore different integration methods for the graph node and edge embeddings and put forward a new IoU (Intersection over Union) guided function, which improves long term tracking and handles identity switches. Second, we introduce a hierarchical sampling strategy to construct sparser graphs which allows to focus the training on more difficult samples. Experimental results demonstrate that a simple online MPN with these two contributions can perform better than many state-of-the-art methods. In addition, our association method generalizes well and can also improve the results of private detection based methods.