MOTS: Multiple Object Tracking for General Categories Based On Few-Shot Method

TL;DR

MOTS introduces a novel multi-object tracking system based on few-shot learning that generalizes across categories, utilizing a two-stage process with adaptive matching and fine matching for unmatched targets.

Contribution

The paper presents MOTS, a new multi-object tracking framework that extends beyond specific categories using a few-shot approach and a two-stage matching process.

Findings

Achieves 88.76% assignment accuracy on MOT16 training set.

Successfully matches 31 categories, including unseen ones.

Demonstrates effective generalization to new categories.

Abstract

Most modern Multi-Object Tracking (MOT) systems typically apply REID-based paradigm to hold a balance between computational efficiency and performance. In the past few years, numerous attempts have been made to perfect the systems. Although they presented favorable performance, they were constrained to track specified category. Drawing on the ideas of few shot method, we pioneered a new multi-target tracking system, named MOTS, which is based on metrics but not limited to track specific category. It contains two stages in series: In the first stage, we design the self-Adaptive-matching module to perform simple targets matching, which can complete 88.76% assignments without sacrificing performance on MOT16 training set. In the second stage, a Fine-match Network was carefully designed for unmatched targets. With a newly built TRACK-REID data-set, the Fine-match Network can perform matching of 31 category targets, even generalizes to unseen categories.