TLDR
测试时自适应 (TTA) 通常假设模型参数可在推理时更新。这一假设对于推理专用加速器、冻结或第三方模型以及内存受限的部署具有较强限制；标准的基于 BatchNorm 的 TTA 配置在不含 BatchNorm 的架构上也可能失效。本文研究模型必须保持冻结时的自适应问题，提出 CASTER，一种无梯度方法，将源类统计量存储在判别性子空间中，基于目标批次矩估计一个类共享的仿射变换，并通过解析方式Test-time adaptation (TTA) typically assumes that model parameters can be updated at inference time. This assumption is restrictive for inference-only accelerators, frozen or third-party models, and memory-constrained deployments, and standard BatchNorm-based TTA configurations may also become inactive on architectures without BatchNorm. We study adaptation when the learned model must remain frozen. We introduce CASTER, a gradient-free method that stores source class statistics in a discriminative subspace, estimates a class-shared affine transformation from target-batch moments, and analytica