Near-Online Multi-target Tracking with Aggregated Local Flow Descriptor

TL;DR

This paper introduces a robust affinity measure called ALFD based on long-term interest point trajectories and a near-online multi-target tracking algorithm that together improve accuracy in challenging datasets.

Contribution

The paper presents a novel ALFD descriptor for better affinity measurement and a near-online tracking algorithm that integrates multiple cues for improved robustness.

Findings

ALFD outperforms conventional affinity metrics in accuracy.

NOMT achieves about 10% higher MOTA on KITTI and MOT datasets.

The combined approach significantly advances multi-target tracking performance.

Abstract

In this paper, we focus on the two key aspects of multiple target tracking problem: 1) designing an accurate affinity measure to associate detections and 2) implementing an efficient and accurate (near) online multiple target tracking algorithm. As the first contribution, we introduce a novel Aggregated Local Flow Descriptor (ALFD) that encodes the relative motion pattern between a pair of temporally distant detections using long term interest point trajectories (IPTs). Leveraging on the IPTs, the ALFD provides a robust affinity measure for estimating the likelihood of matching detections regardless of the application scenarios. As another contribution, we present a Near-Online Multi-target Tracking (NOMT) algorithm. The tracking problem is formulated as a data-association between targets and detections in a temporal window, that is performed repeatedly at every frame. While being efficient, NOMT achieves robustness via integrating multiple cues including ALFD metric, target dynamics, appearance similarity, and long term trajectory regularization into the model. Our ablative analysis verifies the superiority of the ALFD metric over the other conventional affinity metrics. We run a comprehensive experimental evaluation on two challenging tracking datasets, KITTI and MOT datasets. The NOMT method combined with ALFD metric achieves the best accuracy in both datasets with significant margins (about 10% higher MOTA) over the state-of-the-arts.