arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.00996cs.LGstat.ML
收藏

基于动态残差自适应器的潜域学习

Latent Domain Learning with Dynamic Residual Adapters

School of Informatics(信息学院);University of Edinburgh(爱丁堡大学)

Lucas Deecke, Timothy Hospedales, Hakan Bilen

摘要:

深度神经网络的一个实际缺点是它们专用于单个任务和域。虽然领域自适应和多领域学习中的最新技术能够学习更多领域不可知的特征,但它们的成功依赖于领域标签的存在,通常需要手动注释和仔细管理数据集。在这里,我们关注一个探索较少但更现实的案例:从多个领域的数据中学习,而无需访问领域注释。在这种情况下,标准模型训练会导致大型域的过拟合,而忽略较小的域。我们通过动态残余适配器,一个自适应门控机制,有助于考虑潜在的域,再加上最近的风格转移技术的启发增强策略来解决这个限制。我们所提出的方法是检查包含多个潜在域的图像分类任务,我们展示了它的能力,在这些获得强大的性能。动态剩余适配器的性能明显优于具有更大容量的现成网络,并且可以以端到端的方式与现有架构无缝集成。

英文摘要:

A practical shortcoming of deep neural networks is their specialization to a single task and domain. While recent techniques in domain adaptation and multi-domain learning enable the learning of more domain-agnostic features, their success relies on the presence of domain labels, typically requiring manual annotation and careful curation of datasets. Here we focus on a less explored, but more realistic case: learning from data from multiple domains, without access to domain annotations. In this scenario, standard model training leads to the overfitting of large domains, while disregarding smaller ones. We address this limitation via dynamic residual adapters, an adaptive gating mechanism that helps account for latent domains, coupled with an augmentation strategy inspired by recent style transfer techniques. Our proposed approach is examined on image classification tasks containing multiple latent domains, and we showcase its ability to obtain robust performance across these. Dynamic residual adapters significantly outperform off-the-shelf networks with much larger capacity, and can be incorporated seamlessly with existing architectures in an end-to-end manner.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑