弥合领域专业知识与泛化能力以实现性能估计
Bridging Domain Expertise and Generalization for Performance Estimation
在无标签测试集上估计模型真实性能时，可借用外部基础模型的预测作为参考信号，通过温度缩放校准对齐分布后再做置信度加权融合，缓解分布偏移下模型自身输出偏差被放大的问题；这套参考分布构造思路可直接迁移到视觉模型的部署前评估与模型选择。
原文依据 FRAP aligns the prediction distribution of the foundation model with that of the base model by applying temperature-scaled calibration that minimizes their divergence.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论分布偏移下的性能估计旨在预测模型在分布与训练数据不同的无标签测试集上的表现，这一场景需要能够在没有真实标签的情况下可靠反映模型行为的指标。现有方法仅依赖给定模型的输出，而一旦分布发生偏移，这些输出中的偏差会被放大，从而削弱其与真实性能的相关性。受此局限性的启发，我们提出融合参考对齐预测（FRAP），它利用外部基础模型与基座模型的互补优势，构建更可靠的真实标签替代信号。FRAP 通过施加最小化二者散度的温度缩放校准，将基础模型的预测分布与基座模型的预测分布对齐。对齐后的预测通过基于置信度的加权进行融合，形成一个精炼的参考分布，该分布同时整合了基础模型的鲁棒性与基座模型的领域专业知识，并通过度量基座模型预测与该参考分布的一致程度来获得性能估计。在多种数据集与架构上的大量实验表明，在分布偏移下，FRAP 相较具有代表性的性能估计方法取得了持续且显著的提升。
英文原文摘要 1402 字符
Performance estimation under distribution shift aims to predict how a model behaves on an unlabeled test set whose distribution differs from the training data, a scenario that requires reliable indicators that can faithfully reflect model behavior without ground-truth labels. Existing approaches rely solely on the outputs of the given model whose biases are amplified once the distribution shifts, weakening the correlation with the true performance. Motivated by this limitation, we propose Fused Reference Alignment Prediction (FRAP), which leverages the complementary strengths of an external foundation model and the base model to construct a more reliable surrogate of the ground-truth labels. FRAP aligns the prediction distribution of the foundation model with that of the base model by applying temperature-scaled calibration that minimizes their divergence. The aligned predictions are fused through confidence-based weighting into a refined reference distribution that integrates robustness from the foundation model and domain-specific expertise from the base model, and performance estimation is obtained by measuring how closely the base model predictions agree with this reference. Extensive experiments across diverse datasets and architectures show that FRAP provides consistent and substantial improvements over representative performance-estimation methods under distribution shift.