RICASSO：基于类别感知自监督离群值暴露的强化不平衡学习
RICASSO: Reinforced Imbalance Learning with Class-Aware Self-Supervised Outliers Exposure
AI总结：
提出RICASSO框架，通过数据混合生成伪OOD数据，结合三类策略同时解决长尾识别与OOD检测问题，在iNaturalist2018上达到SOTA并显著超越基线。
摘要：
在现实场景中，深度学习模型经常面临不平衡（长尾）数据和分布外（OOD）数据的双重挑战。然而，现有的联合方法依赖真实的OOD数据，这导致了不必要的权衡。相反，我们的研究表明，数据混合（data mixing）作为一种用于长尾识别的强大增强技术，可以生成同时具备分布内（ID）数据和OOD数据特征的伪OOD数据。因此，通过使用混合数据替代真实OOD数据，我们可以整体性地解决长尾识别和OOD检测问题。我们提出了一个名为Reinforced Imbalance Learning with Class-Aware Self-Supervised Outliers Exposure（RICASSO）的统一框架，其中“自监督”表示我们仅使用ID数据进行离群值暴露。RICASSO包含三个主要策略：1. 基于范数-奇数对偶性的离群值暴露：使用混合数据作为伪OOD数据，通过单一损失函数同时实现ID数据重平衡和离群值暴露。2. 模糊感知的Logits调整：利用ID数据的模糊性自适应地重新校准logits。3. 对比边界-中心学习：结合虚拟边界学习和双熵中心学习，利用混合数据实现更好的特征分离和聚类，并通过表征一致性学习增强鲁棒性。大量实验表明，RICASSO在长尾识别中达到了最先进的性能，并且与基线相比显著改善了OOD检测（在iNaturalist2018数据集上AUROC提升27%，FPR降低61%）。在iNaturalist2018上，我们甚至优于使用真实OOD数据的方法。代码将很快公开。
英文摘要：
In real-world scenarios, deep learning models often face challenges from both imbalanced (long-tailed) and out-of-distribution (OOD) data. However, existing joint methods rely on real OOD data, which leads to unnecessary trade-offs. In contrast, our research shows that data mixing, a potent augmentation technique for long-tailed recognition, can generate pseudo-OOD data that exhibit the features of both in-distribution (ID) data and OOD data. Therefore, by using mixed data instead of real OOD data, we can address long-tailed recognition and OOD detection holistically. We propose a unified framework called Reinforced Imbalance Learning with Class-Aware Self-Supervised Outliers Exposure (RICASSO), where "self-supervised" denotes that we only use ID data for outlier exposure. RICASSO includes three main strategies: Norm-Odd-Duality-Based Outlier Exposure: Uses mixed data as pseudo-OOD data, enabling simultaneous ID data rebalancing and outlier exposure through a single loss function. Ambiguity-Aware Logits Adjustment: Utilizes the ambiguity of ID data to adaptively recalibrate logits. Contrastive Boundary-Center Learning: Combines Virtual Boundary Learning and Dual-Entropy Center Learning to use mixed data for better feature separation and clustering, with Representation Consistency Learning for robustness. Extensive experiments demonstrate that RICASSO achieves state-of-the-art performance in long-tailed recognition and significantly improves OOD detection compared to our baseline (27% improvement in AUROC and 61% reduction in FPR on the iNaturalist2018 dataset). On iNaturalist2018, we even outperforms methods using real OOD data. The code will be made public soon.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
