发表机构
Institute of Business Administration Karachi（卡拉奇工商管理学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出零训练Fisher几何对齐（ZFGA），利用信息几何对齐特征空间，无需训练或参数修改，在多个模型和数据集上提升分布偏移鲁棒性，且不损害任何模型性能。
AI 中文摘要
深度视觉模型在分布偏移下常出现性能退化。测试时自适应可提升鲁棒性，但通常需要迭代优化、超参数调优及多次前向-反向传播。我们提出零训练Fisher几何对齐（ZFGA），一种闭式方法，在不修改模型参数的情况下提升协变量偏移下的鲁棒性。ZFGA基于分布偏移扭曲特征空间几何的观察。它估计预测分布关于特征嵌入的Fisher信息矩阵，并应用线性变换，使测试特征的Fisher几何与从干净数据计算的参考几何对齐。这提供了特征空间中自然梯度启发的预处理步骤。我们在CIFAR-10-C和ImageNet-C上使用ResNet-50、DINO ViT-S/16及CLIP ViT-B/32评估ZFGA。ZFGA在所有三个模型上一致优于零样本推理，尽管并非每个模型的最强方法。协方差白化在ResNet-50上表现更好，而Fisher白化与ZFGA在CLIP上统计上无显著差异。在六种免训练和基于梯度的方法（协方差白化、Fisher白化、TENT、T3A、LAME和AdaNPC）中，ZFGA是唯一未对三个模型家族中任何一个造成实质性损害的方法。Fisher几何失真与ZFGA增益正相关（Pearson r = 0.366，p = 0.017），初步证据表明几何错位导致鲁棒性退化。ZFGA推理时仅需前向传播和矩阵运算，为基于优化的测试时自适应提供了轻量且确定性的替代方案。
英文摘要
Deep vision models often degrade under distribution shift. Test-time adaptation can improve robustness but typically requires iterative optimization, hyperparameter tuning, and multiple forward-backward passes. We propose Zero-Training Fisher Geometry Alignment (ZFGA), a closed-form method that improves robustness under covariate shift without modifying model parameters. ZFGA is based on the observation that distribution shifts distort feature-space geometry. It estimates the Fisher information matrix of the predictive distribution with respect to feature embeddings and applies a linear transformation that aligns test-feature Fisher geometry with a reference geometry computed from clean data. This provides a natural-gradient-inspired preconditioning step in feature space. We evaluate ZFGA on CIFAR-10-C and ImageNet-C using ResNet-50, DINO ViT-S/16, and CLIP ViT-B/32. ZFGA consistently improves over zero-shot inference across all three models, although it is not the strongest method for every model. Covariance whitening performs better on ResNet-50, while Fisher whitening is statistically indistinguishable from ZFGA on CLIP. Across six training-free and gradient-based alternatives (covariance whitening, Fisher whitening, TENT, T3A, LAME, and AdaNPC), ZFGA is the only method that does not substantially harm any of the three model families. The Fisher geometry distortion is also positively correlated with ZFGA gain (Pearson r = 0.366, p = 0.017), providing preliminary evidence that geometric misalignment contributes to robustness degradation. ZFGA requires only forward passes and matrix operations at inference time, offering a lightweight and deterministic alternative to optimization-based test-time adaptation.
