UnsMOT: Unified Framework for Unsupervised Multi-Object Tracking with Geometric Topology Guidance

TL;DR

This paper introduces UnsMOT, an unsupervised multi-object tracking framework that combines appearance, motion, and geometric information to improve tracking accuracy without requiring annotated data.

Contribution

The novel UnsMOT framework integrates appearance, motion, and geometric features using CNN, RNN, and GNN models for unsupervised multi-object tracking.

Findings

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

Effectively combines appearance, motion, and geometric cues.

Improves tracking accuracy without labeled training data.

Abstract

Object detection has long been a topic of high interest in computer vision literature. Motivated by the fact that annotating data for the multi-object tracking (MOT) problem is immensely expensive, recent studies have turned their attention to the unsupervised learning setting. In this paper, we push forward the state-of-the-art performance of unsupervised MOT methods by proposing UnsMOT, a novel framework that explicitly combines the appearance and motion features of objects with geometric information to provide more accurate tracking. Specifically, we first extract the appearance and motion features using CNN and RNN models, respectively. Then, we construct a graph of objects based on their relative distances in a frame, which is fed into a GNN model together with CNN features to output geometric embedding of objects optimized using an unsupervised loss function. Finally, associations between objects are found by matching not only similar extracted features but also geometric embedding of detections and tracklets. Experimental results show remarkable performance in terms of HOTA, IDF1, and MOTA metrics in comparison with state-of-the-art methods.