Unsupervised domain adaptive re-identification: Theory and practice
作者：
Highlights：
• We introduce the theoretical guarantees of unsupervised domain adaptive re-ID based on [2]. A DA-learnability result is shown under three assumptions that concerning the feature space. To the best of our knowledge, our paper is the first theoretical analysis work on domain adaptive re-ID tasks.
• We theoretically turn the goal of satisfying the assumptions into tractable loss functions on the encoder network and data samples.
• A self-training scheme is proposed to iteratively minimizing the loss functions. Our framework is applicable to all re-ID tasks and the effectiveness is verified on large-scale datasets for diverse re-ID tasks.
摘要
•We introduce the theoretical guarantees of unsupervised domain adaptive re-ID based on [2]. A DA-learnability result is shown under three assumptions that concerning the feature space. To the best of our knowledge, our paper is the first theoretical analysis work on domain adaptive re-ID tasks.•We theoretically turn the goal of satisfying the assumptions into tractable loss functions on the encoder network and data samples.•A self-training scheme is proposed to iteratively minimizing the loss functions. Our framework is applicable to all re-ID tasks and the effectiveness is verified on large-scale datasets for diverse re-ID tasks.
论文关键词：Person re-identification,Unsupervised domain adaptation
论文评审过程：Received 24 April 2019, Revised 26 November 2019, Accepted 15 December 2019, Available online 10 January 2020, Version of Record 26 February 2020.
论文官网地址：https://doi.org/10.1016/j.patcog.2019.107173
