TransMOT: Spatial-Temporal Graph Transformer for Multiple Object Tracking

TL;DR

TransMOT introduces a graph transformer-based approach for multiple object tracking that models spatial-temporal interactions efficiently, achieving state-of-the-art accuracy while reducing computational costs.

Contribution

The paper presents a novel graph transformer architecture for MOT, incorporating sparse graphs and a cascade association framework for improved speed and accuracy.

Findings

Achieves state-of-the-art performance on MOT datasets.

More computationally efficient than traditional Transformers.

Effectively handles low-score detections and occlusions.

Abstract

Tracking multiple objects in videos relies on modeling the spatial-temporal interactions of the objects. In this paper, we propose a solution named TransMOT, which leverages powerful graph transformers to efficiently model the spatial and temporal interactions among the objects. TransMOT effectively models the interactions of a large number of objects by arranging the trajectories of the tracked objects as a set of sparse weighted graphs, and constructing a spatial graph transformer encoder layer, a temporal transformer encoder layer, and a spatial graph transformer decoder layer based on the graphs. TransMOT is not only more computationally efficient than the traditional Transformer, but it also achieves better tracking accuracy. To further improve the tracking speed and accuracy, we propose a cascade association framework to handle low-score detections and long-term occlusions that require large computational resources to model in TransMOT. The proposed method is evaluated on multiple benchmark datasets including MOT15, MOT16, MOT17, and MOT20, and it achieves state-of-the-art performance on all the datasets.