为何不采用超参数友好的优化？一种用于长尾识别的单调自适应范数重缩放方法
Why Not Hyperparameter-Friendly Optimisation? A Monotonic Adaptive Norm Rescaling Approach For Long-Tailed Recognition
SAMN 用池相邻违例算法直接对每类权重范数施加单调性约束，替代了需要调参的参数正则化，为长尾识别中的分类器重训练提供了一个无需超参数的通用后处理模块，可即插即用地集成到现有解耦框架中。
原文依据 SAMN avoids the need for parameter regularization. It directly enforces monotonicity on per-class weight norms using the Pool Adjacent Violators Algorithm
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论长尾识别对深度学习构成了重大挑战。两阶段解耦范式将表征学习与分类器重训练分离，提供了一种有前景的解决方案。在分类器重训练阶段，自适应范数重缩放是一种流行技术。它通过参数正则化调整每类权重范数，这不可避免地引入了超参数。然而，许多研究报告称长尾识别对这些超参数敏感，因为它们的设置会显著影响性能。本文首先从类别条件分布的角度为范数重缩放方法提供支持。此外，我们提出了一种简单而有效的方法，称为自自适应单调归一化（SAMN）。SAMN 避免了参数正则化的需要。它使用池相邻违例算法直接对每类权重范数施加单调性，使该方法对超参数友好。SAMN 是一种通用策略，可与其他方法无缝集成以提升性能。在基准数据集上的实验表明，我们的方法显著提升了长尾识别性能，通常达到最优结果。代码可在 https://github.com/Zhangshuojackpot/SAMN 获取。
英文原文摘要 1300 字符
Long-tailed recognition poses a significant challenge for deep learning. The two-stage decoupling paradigm, which separates representation learning from classifier retraining, offers a promising solution. During the classifier retraining stage, adaptive norm rescaling is a popular technique. It adjusts the per-class weight norms via parameter regularization, which inevitably introduces hyperparameters. However, many studies report that long-tailed recognition is sensitive to these hyperparameters, as their setup significantly impacts performance. In this paper, we first provide a class-conditional distribution perspective to support norm rescaling methods. Furthermore, we propose a simple but effective approach called Self-Adaptive Monotonic Normalization (SAMN). SAMN avoids the need for parameter regularization. It directly enforces monotonicity on per-class weight norms using the Pool Adjacent Violators Algorithm, making the method hyperparameter-friendly. SAMN is a universal strategy that integrates seamlessly with other methods for enhanced performance. Experiments on benchmark datasets demonstrate that our method significantly boosts long-tailed recognition performance, often achieving state-of-the-art results. Codes are available at https://github.com/Zhangshuojackpot/SAMN.