发表机构
Concordia University; University of Toronto; Mila – Quebec AI Institute（康考迪亚大学; 多伦多大学; 米拉-魁北克人工智能研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究少样本测试时的域适应问题，提出DA-MergeLoRA框架，将LoRA微调与模型合并相结合，通过元学习训练超网络生成合并因子，在多个域适应数据集上取得最优性能。
AI 中文摘要
少样本测试时的域适应（FSTT-DA）旨在仅使用少量未标记的目标样本使模型适应新领域。此设置比典型的域适应设置更现实，后者假设在源训练期间可访问目标数据。然而，先前的FSTT-DA方法无法有效利用源域特定知识，依赖浅批量归一化更新、将模型视为黑箱的基于提示的方法或无法捕捉跨域关系的集成策略。为解决这些限制，我们引入了一个新的FSTT-DA框架，将LoRA微调与模型合并相结合。在我们的方法中，针对每个源域在CLIP的视觉编码器上对单独的LoRA模块进行微调。由于LoRA仅修改模型参数的一小部分，它在内部学习域特定特征时保留了基础模型的广义知识。为使学到的知识适应特定目标域，我们提出一个通过元学习训练的超网络，生成逐列合并因子以组合LoRA模块。给定一小批目标图像，超网络产生合并权重，将源LoRA模块融合成单个适应表示。我们的结果展示了在各种域适应数据集上的最优性能。我们的代码可在该https URL公开获取。
英文摘要
Few-shot Test-Time Domain Adaptation (FSTT-DA) seeks to adapt models to novel domains using only a handful of unlabeled target samples. This setting is more realistic than typical domain adaptation setups, which assume access to target data during source training. However, prior FSTT-DA approaches fail to effectively leverage source domain-specific knowledge, relying on shallow batch normalization updates, prompt-based methods that treat the model as a black box, or ensembling strategies that do not capture cross-domain relationships. To address these limitations, we introduce a new FSTT-DA framework that integrates LoRA fine-tuning with model merging. In our approach, separate LoRA modules are fine-tuned on CLIP's vision encoder for each source domain. Since LoRA modifies only a small fraction of the model's parameters, it retains the base model's generalized knowledge while internally learning domain-specific features. To adapt the learned knowledge to a specific target domain, we propose a hypernetwork trained via meta-learning that generates per-column merging factors to combine LoRA modules. Given a small batch of target images, the hypernetwork produces merging weights that fuse source LoRA modules into a single adapted representation. Our results demonstrate state-of-the-art performance across various domain adaptation datasets. Our code is publicly available at https://github.com/nahbois4321/DA-MergeLoRA.
CommentsECCV 2026, Code: https://github.com/nahbois4321/DA-MergeLoRA
