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

2025-08-20

AI Summary

To address label shift, covariate shift, and emergent unknown classes under severe pretraining class imbalance in open-world settings, this paper proposes an imbalance-aware continual adaptation framework. Methodologically, we introduce a class-balanced contrastive pretraining strategy, coupled with dynamic pseudo-label generation and a selective parameter activation mechanism that updates only critical parameters during fine-tuning—substantially reducing computational overhead. Theoretical analysis and empirical evaluation demonstrate robust representation learning for both minority and unknown classes. On multiple open-world benchmarks, our approach surpasses state-of-the-art methods in both classification accuracy (especially +12.3% on minority classes) and inference efficiency (37% reduction in FLOPs). To the best of our knowledge, this is the first work to achieve efficient and robust continual learning under imbalanced pretraining conditions.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Web Mining and Content Analysis: Large pretrained models with web data

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.

Problem

• Handles open-world problems with imbalanced pre-training data
• Improves classification for underrepresented classes using contrastive learning
• Generates reliable pseudo-labels to enhance model robustness

Innovation

• Contrastive pre-training for imbalanced data
• Pseudo-label generation for model robustness
• Selective activation to reduce computation