A Bottom-Up Clustering Approach to Unsupervised Person Re-Identification

By Yutian Lin, Xuanyi Dong, Liang Zheng, Yan Yan, Yi Yang. AAAI Conference on Artificial Intelligence, 2019

583 citations recorded by the source corpus.

Most person re-identification (re-ID) approaches are based on supervised learning, which requires intensive manual annotation for training data. However, it is not only resourceintensive to acquire identity annotation but also impractical to label the large-scale real-world data. To relieve this problem, we propose a bottom-up clustering (BUC) approach to jointly optimize a convolutional neural network (CNN) and the relationship among the individual samples. Our algorithm considers two fundamental facts in the re-ID task, i.e., diversity across different identities and similarity within the same identity. Specifically, our algorithm starts with regarding individual sample as a different identity, which maximizes the diversity over each identity. Then it gradually groups similar samples into one identity, which increases the similarity within each identity. We utilizes a diversity regularization term in the bottom-up clustering procedure to balance the data volume of each cluster. Finally, the model achieves an effective trade-off between the diversity and similarity. We conduct extensive experiments on the large-scale image and video re-ID datasets, including Market-1501, DukeMTMCreID, MARS and DukeMTMC-VideoReID. The experimental results demonstrate that our algorithm is not only superior to state-of-the-art unsupervised re-ID approaches, but also performs favorably than competing transfer learning and semi-supervised learning methods.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/123478b496a3fa39a9043ccaa660e81c473a14e9

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Cluster Contrast for Unsupervised Person Re-Identification (2021, background): “[24] treats each individual sample as a cluster, and then gradually groups similar samples into one cluster to generate pseudo labels.”
Hierarchical Clustering With Hard-Batch Triplet Loss for Person Re-Identification (2019, methodology): “For all clustering-guided re-ID [14, 6, 20], the quality of pseudo labels generated by clusters directly determines the performance of the model.”
Lifelong Person Re-Identification via Adaptive Knowledge Accumulation (2021, background): “1, previous person ReID works are performed in four different setups: 1) Fully-supervised (FS) methods investigate and exploit different network structures and loss functions [53, 31, 47, 30]; 2) Unsupervised domain adaption (UDA) is introduced to mitigate the domain gaps between source and target d”
Identity-Guided Human Semantic Parsing for Person Re-Identification (2020, methodology): “[26] also use clustering to solve the unsupervised person re-ID task.”
Unsupervised Person Re-Identification via Multi-Label Classification (2020, methodology): “Compared with previous methods [20, 34], which classify each image into a single class, the multi-label classification has potential to exhibit better efficiency and accuracy.”
Improving Person Re-identification by Attribute and Identity Learning (2017, methodology): “Recently, some semi-supervised methods [37, 19] and unsupervised methods [38, 22] has been proposed to address the data problem for re-ID.”
Patch-based Knowledge Distillation for Lifelong Person Re-Identification (2022, background): “Person ReID has made remarkable progress under various settings such as supervised learning [18, 40, 57], unsupervised learning [9, 36] and unsupervised domain adaptation [12, 15, 66, 75, 76].”
Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification (2020, methodology): “1 CLUSTERING-BASED UDA METHODS REVISIT State-of-the-art UDA methods (Fan et al., 2018; Lin et al., 2019; Zhang et al., 2019b; Yang et al., 2019) follow a similar general pipeline.”
Pose-Guided Feature Alignment for Occluded Person Re-Identification (2019, background): “Recent years, deep learning methods [39, 34, 13, 27, 3, 25, 7, 6, 36, 21] for person re-id show signiﬁcant superiority on retrieval accuracy.”
Comprehensive survey on hierarchical clustering algorithms and the recent developments (2022, methodology): “The methods which combine traditional hierarchical clustering with CNN or the hierarchical clustering based on deep learning techniques have been proposed recently (Zeng et al. 2020; Zhao et al. 2020; Lin et al. 2019).”