Deep Domain Confusion: Maximizing for Domain Invariance
Eric Tzeng, Judy Hoffman, Ning Zhang UC Berkeley, EECS、ICSI Kate Saenko UMass Lowell, CS Trevor Darrell UC Berkeley, EECS、ICSI
Abstract
Recent reports suggest that a generic supervised deep CNN model trained on a large-scale dataset reduces, but does not remove, dataset bias on a standard benchmark. Fine-tuning deep models in a new domain can require a significant amount of data, which for many applications is simply not available. We propose a new CNN architecture which introduces an adaptation layer and an additional domain confusion loss, to learn a representation that is both semantically meaningful and domain invariant. We additionally show that a domain confusion metric can be used for model selection to determine the dimension of an adaptation layer and the best position for the layer in the CNN architecture. Our proposed adaptation method offers empirical performance which exceeds previously published results on a standard benchmark visual domain adaptation task.
中文速览
视觉识别模型在换了拍摄环境或设备后往往表现大幅下降，根本原因是训练域与测试域的数据分布存在偏差（domain bias）。针对这一问题，研究者在经典深度卷积神经网络（CNN）架构中插入一个低维"瓶颈"适配层，并在该层上同时施加分类损失和基于最大均值差异（Maximum Mean Discrepancy, MMD）的域混淆损失，迫使网络学到既能区分类别、又对域差异不敏感的特征表示。此外，他们还利用MMD指标自动决定适配层插入的位置和维度，无需人工调参。在Office标准跨域基准上，该方法在有标注和无标注目标域两种设定下均超越了此前所有已发表方法，在变化较小的场景中目标域准确率甚至达到96%。这项工作的意义在于，它提供了一种系统化、可自动调整的深度域适应框架，为样本稀少的实际应用场景下的跨域视觉识别提供了切实可行的解决思路。
原文 arXiv:1412.3474;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1412.3474v1