跨域小样本分类的无训练谱转导精化
Training-Free Spectral Transductive Refinement for Cross-Domain Few-Shot Classification
- Islamic University of Technology（伊斯兰理工大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
提出无训练谱转导精化（STR），利用谱几何和伪标注迭代精化，在跨域小样本分类中仅推理时提升性能，1-shot平均最优。
AI中文摘要：
在域偏移和单样本监督下，使用冻结视觉特征的小样本识别尤其脆弱，此时单个标注图像对其类别的估计不可靠。我们探究在仅测试时、不重新训练编码器或增强源域的情况下，这种脆弱性能在多大程度上被降低。我们提出谱转导精化（STR），一种无训练的转导推理规则，利用完整支持-查询片段（episode）的几何结构。给定冻结嵌入，STR构建联合k近邻图，将片段映射到归一化拉普拉斯谱坐标系，从标注支持集初始化类代表，并使用伪标注查询迭代精化它们。我们在两种协议下评估STR。使用冻结ResNet-18特征的受控组件研究表明，谱精化在五个偏移域上持续优于单原型谱初始化，在支持估计最弱的单样本设置中增益最大。然后，我们使用标准miniImageNet预训练ResNet-10骨干在八个既定目标域上，将STR与近期跨域小样本学习（CD-FSL）方法进行基准比较。完全在推理时运行，STR在比较方法中取得最高的1-shot平均准确率，并在5-shot下保持竞争力，与依赖重源域元训练增强的方法相媲美。由于STR是转导的，我们明确报告其设置。诊断将其增益归因于谱坐标中的迭代精化，而非原型容量增加，后者在我们的配置中保持非激活。
英文摘要：
Few-shot recognition with frozen visual features is especially fragile under domain shift and one-shot supervision, where a single labelled image is an unreliable estimate of its class. We ask how far this fragility can be reduced purely at test time, without retraining the encoder or augmenting the source domain. We present Spectral Transductive Refinement (STR), a training-free transductive inference rule that exploits the geometry of the complete support-query episode. Given frozen embeddings, STR builds a joint k-nearest-neighbour graph, maps the episode into a normalized-Laplacian spectral coordinate system, initializes class representatives from the labelled support, and iteratively refines them using pseudo-labelled queries. We evaluate STR under two protocols. A controlled component study with frozen ResNet-18 features shows that spectral refinement consistently improves over single-prototype spectral initialization across five shifted domains, with the largest gains in the one-shot regime where support estimates are weakest. We then benchmark STR against recent Cross-Domain Few-Shot Learning (CD-FSL) methods using the standard miniImageNet-pretrained ResNet-10 backbone over eight established target domains. Operating entirely at inference time, STR attains the highest 1-shot average among compared methods and remains competitive at 5-shot, rivalling approaches relying on heavy source-domain meta-training augmentations. Because STR is transductive, we report its setting explicitly. Diagnostics attribute its gains to iterative refinement in spectral coordinates rather than added prototype capacity, which remains inactive in our configuration.
