用于可泛化行人重识别的自适应域特定归一化
摘要
尽管现有的行人重识别(Re-ID)方法已展现出令人印象深刻的准确率,但它们大多在未见目标域上泛化性差。因此,可泛化行人重识别近期受到越来越多的关注,其在源域上训练模型,使其无需更新模型即可在未见目标域上良好泛化。本文中,我们提出一种用于可泛化行人重识别的新型自适应域特定归一化方法(AdsNorm)。它将未见目标域描述为已知源域的组合,并通过元学习流程显式地学习具有目标分布的域特定表示,以提升模型的泛化能力。具体而言,AdsNorm 利用批归一化层收集各个源域的特征,并使用这些特征将源域映射到一个共享潜空间,其中域相关性由不同域特定归一化统计量与特征的距离函数度量。在测试阶段,AdsNorm 将来自未见目标域的图像投影到同一潜空间,并通过域相关性自适应地整合携带源分布的域特定特征,以在未见目标域上学习更具泛化性的聚合表示。考虑到训练时目标域不可用,我们提出一种结合定制关系损失的元学习算法来优化一个有效且高效的集成模型。大量实验表明 AdsNorm 优于最先进的方法。代码见:https://github.com/hzphzp/AdsNorm。
引用
@article{arxiv.2105.03042,
title = {Adaptive Domain-Specific Normalization for Generalizable Person Re-Identification},
author = {Jiawei Liu and Zhipeng Huang and Kecheng Zheng and Dong Liu and Xiaoyan Sun and Zheng-Jun Zha},
journal= {arXiv preprint arXiv:2105.03042},
year = {2021}
}
备注
Withdraw this paper for internal review. Since we were not familiar with the use of arXiv, our initial manuscript was uploaded by mistake and we found many inappropriate and unmodified parts of it (such as the experimental results in Table 2,3, the Equation 13). I am sorry to say that this work still needs to be further completed and we do not intend to use it for publication