DARTH: Holistic Test-time Adaptation for Multiple Object Tracking
This addresses the need for robust MOT in autonomous driving to prevent life-critical failures, representing a novel solution to an unsolved adaptation problem.
The paper tackles the problem of domain shift in multiple object tracking (MOT) for autonomous driving by introducing DARTH, a holistic test-time adaptation framework, which improves source model performance across various domain shifts such as sim-to-real and outdoor-to-indoor.
Multiple object tracking (MOT) is a fundamental component of perception systems for autonomous driving, and its robustness to unseen conditions is a requirement to avoid life-critical failures. Despite the urge of safety in driving systems, no solution to the MOT adaptation problem to domain shift in test-time conditions has ever been proposed. However, the nature of a MOT system is manifold - requiring object detection and instance association - and adapting all its components is non-trivial. In this paper, we analyze the effect of domain shift on appearance-based trackers, and introduce DARTH, a holistic test-time adaptation framework for MOT. We propose a detection consistency formulation to adapt object detection in a self-supervised fashion, while adapting the instance appearance representations via our novel patch contrastive loss. We evaluate our method on a variety of domain shifts - including sim-to-real, outdoor-to-indoor, indoor-to-outdoor - and substantially improve the source model performance on all metrics. Code: https://github.com/mattiasegu/darth.