MOANA: An Online Learned Adaptive Appearance Model for Robust Multiple Object Tracking in 3D

TL;DR

This paper introduces MOANA, an online adaptive appearance model for robust 3D multiple object tracking that effectively handles occlusion, appearance changes, and incorporates 3D geometry, outperforming state-of-the-art methods in real-time.

Contribution

The paper presents a novel online adaptive appearance model that encodes long-term appearance changes and integrates 3D geometry for improved multi-object tracking.

Findings

Outperforms state-of-the-art on MOTChallenge 3D benchmark

Achieves real-time processing on standard desktop CPU

Demonstrates superior performance on 2D MOTChallenge benchmark

Abstract

Multiple object tracking has been a challenging field, mainly due to noisy detection sets and identity switch caused by occlusion and similar appearance among nearby targets. Previous works rely on appearance models built on individual or several selected frames for the comparison of features, but they cannot encode long-term appearance changes caused by pose, viewing angle and lighting conditions. In this work, we propose an adaptive model that learns online a relatively long-term appearance change of each target. The proposed model is compatible with any feature of fixed dimension or their combination, whose learning rates are dynamically controlled by adaptive update and spatial weighting schemes. To handle occlusion and nearby objects sharing similar appearance, we also design cross-matching and re-identification schemes based on the application of the proposed adaptive appearance models. Additionally, the 3D geometry information is effectively incorporated in our formulation for data association. The proposed method outperforms all the state-of-the-art on the MOTChallenge 3D benchmark and achieves real-time computation with only a standard desktop CPU. It has also shown superior performance over the state-of-the-art on the 2D benchmark of MOTChallenge.