Self-ensembling for visual domain adaptation
French, G. Mackiewicz, M. Fisher, M.
Abstract
This paper explores the use of self-ensembling for visual domain adaptation problems. Our technique is derived from the mean teacher variant [29] of temporal ensembling [14], a technique that achieved state of the art results in the area of semi-supervised learning. We introduce a number of modifications to their approach for challenging domain adaptation scenarios and evaluate its effectiveness. Our approach achieves state of the art results in a variety of benchmarks, including our winning entry in the VISDA-2017 visual domain adaptation challenge. In small image benchmarks, our algorithm not only outperforms prior art, but can also achieve accuracy that is close to that of a classifier trained in a supervised fashion.
中文速览
视觉域自适应（visual domain adaptation）的核心难题在于：用带标注的源域数据训练出的模型，在没有标注的目标域上表现往往大幅下滑。本文将半监督学习领域表现卓越的"均值教师"（mean teacher）自集成方法引入域自适应场景，并针对跨域挑战引入了置信度阈值过滤（confidence thresholding）和类别平衡（class balancing）等关键改进，同时让源域与目标域各自使用独立的批归一化统计量，从而缓解两域分布差异带来的干扰。在多个小图像基准测试上，该方法不仅超越了此前所有方法，部分结果甚至接近有监督训练的精度上限；在大规模合成到真实场景的 VisDA-2017 挑战赛中，该方案赢得了冠军。这项工作表明，无需复杂的对抗生成网络，仅凭自集成与简洁的训练策略即可在域自适应任务上达到顶尖水平，为实际应用中利用廉价合成标注数据迁移到真实场景提供了高效可行的解决路径。
原文 arXiv:1706.05208;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1706.05208v4