Open-World Test-Time Training: Self-Training with Contrast Learning

TL;DR

This paper introduces Open-World Test-Time Training with a novel contrastive learning approach called OWDCL, which improves model robustness and performance in open-world scenarios with unknown and strong OOD data.

Contribution

The paper proposes OWDCL, a contrastive learning method that enhances early feature extraction and robustness in open-world TTT, addressing limitations of existing methods.

Findings

OWDCL outperforms existing TTT methods on benchmark datasets.

Contrastive learning improves early-stage feature extraction.

Model robustness is significantly enhanced in open-world scenarios.

Abstract

Traditional test-time training (TTT) methods, while addressing domain shifts, often assume a consistent class set, limiting their applicability in real-world scenarios characterized by infinite variety. Open-World Test-Time Training (OWTTT) addresses the challenge of generalizing deep learning models to unknown target domain distributions, especially in the presence of strong Out-of-Distribution (OOD) data. Existing TTT methods often struggle to maintain performance when confronted with strong OOD data. In OWTTT, the focus has predominantly been on distinguishing between overall strong and weak OOD data. However, during the early stages of TTT, initial feature extraction is hampered by interference from strong OOD and corruptions, resulting in diminished contrast and premature classification of certain classes as strong OOD. To address this, we introduce Open World Dynamic Contrastive Learning (OWDCL), an innovative approach that utilizes contrastive learning to augment positive sample pairs. This strategy not only bolsters contrast in the early stages but also significantly enhances model robustness in subsequent stages. In comparison datasets, our OWDCL model has produced the most advanced performance.