OASIS: Open-world Adaptive Self-supervised and Imbalanced-aware System

TL;DR

This paper introduces a contrastive pre-training and post-training framework that improves open-world classification, especially for imbalanced datasets, by generating reliable pseudo-labels and optimizing adaptation processes.

Contribution

It presents a novel contrastive-based pre-training and post-training method that effectively handles class imbalance and enhances model robustness in open-world environments.

Findings

Outperforms state-of-the-art adaptation techniques in accuracy

Demonstrates robustness in imbalanced open-world scenarios

Reduces computational overhead with selective activation

Abstract

The expansion of machine learning into dynamic environments presents challenges in handling open-world problems where label shift, covariate shift, and unknown classes emerge. Post-training methods have been explored to address these challenges, adapting models to newly emerging data. However, these methods struggle when the initial pre-training is performed on class-imbalanced datasets, limiting generalization to minority classes. To address this, we propose a method that effectively handles open-world problems even when pre-training is conducted on imbalanced data. Our contrastive-based pre-training approach enhances classification performance, particularly for underrepresented classes. Our post-training mechanism generates reliable pseudo-labels, improving model robustness against open-world problems. We also introduce selective activation criteria to optimize the post-training process, reducing unnecessary computation. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art adaptation techniques in both accuracy and efficiency across diverse open-world scenarios.