IEEE Trans Image Process. 2022;31:3606-3617. doi: 10.1109/TIP.2022.3173163. Epub 2022 May 26.
Unsupervised person re-identification (Re-ID) aims to match pedestrian images from different camera views in an unsupervised setting. Existing methods for unsupervised person Re-ID are usually built upon the pseudo labels from clustering. However, the result of clustering depends heavily on the quality of the learned features, which are overwhelmingly dominated by colors in images. In this paper, we attempt to suppress the negative dominating influence of colors to learn more effective features for unsupervised person Re-ID. Specifically, we propose a Cluster-guided Asymmetric Contrastive Learning (CACL) approach for unsupervised person Re-ID, in which clustering result is leveraged to guide the feature learning in a properly designed asymmetric contrastive learning framework. In CACL, both instance-level and cluster-level contrastive learning are employed to help the siamese network learn discriminant features with respect to the clustering result within and between different data augmentation views, respectively. In addition, we also present a cluster refinement method, and validate that the cluster refinement step helps CACL significantly. Extensive experiments conducted on three benchmark datasets demonstrate the superior performance of our proposal.
无监督行人重识别（Re-ID）旨在在无监督的环境中匹配来自不同摄像机视角的行人图像。现有的无监督行人 Re-ID 方法通常基于聚类的伪标签构建。然而，聚类的结果严重依赖于学习特征的质量，而这些特征主要由图像中的颜色主导。在本文中，我们试图抑制颜色的负面影响，以学习更有效的无监督行人 Re-ID 特征。具体来说，我们提出了一种基于聚类引导的非对称对比学习（CACL）方法用于无监督行人 Re-ID，其中利用聚类结果在适当设计的非对称对比学习框架中指导特征学习。在 CACL 中，同时使用实例级和聚类级对比学习，分别帮助孪生网络在不同数据增强视图内和视图之间基于聚类结果学习具有鉴别力的特征。此外，我们还提出了一种聚类细化方法，并验证了聚类细化步骤有助于 CACL 显著提高性能。在三个基准数据集上进行的广泛实验证明了我们的方法的优越性。
