Unsupervised Person Re-identification by Deep Asymmetric Metric Embedding

Abstract

Person re-identification (Re-ID) aims to match identities across non-overlapping camera views. Researchers have proposed many supervised Re-ID models which require quantities of cross-view pairwise labelled data. This limits their scalabilities to many applications where a large amount of data from multiple disjoint camera views is available but unlabelled. Although some unsupervised Re-ID models have been proposed to address the scalability problem, they often suffer from the view-specific bias problem which is caused by dramatic variances across different camera views, e.g., different illumination, viewpoints and occlusion. The dramatic variances induce specific feature distortions in different camera views, which can be very disturbing in finding cross-view discriminative information for Re-ID in the unsupervised scenarios, since no label information is available to help alleviate the bias. We propose to explicitly address this problem by learning an unsupervised asymmetric distance metric based on cross-view clustering. The asymmetric distance metric allows specific feature transformations for each camera view to tackle the specific feature distortions. We then design a novel unsupervised loss function to embed the asymmetric metric into a deep neural network, and therefore develop a novel unsupervised deep framework named the DEep Clustering-based Asymmetric MEtric Learning (DECAMEL). In such a way, DECAMEL jointly learns the feature representation and the unsupervised asymmetric metric. DECAMEL learns a compact cross-view cluster structure of Re-ID data, and thus help alleviate the view-specific bias and facilitate mining the potential cross-view discriminative information for unsupervised Re-ID. Extensive experiments on seven benchmark datasets whose sizes span several orders show the effectiveness of our framework.

arXiv commentTo appear in TPAMI

References in corpus (9)

• Caffe: Convolutional Architecture for Fast Feature Embedding
• BEGAN: Boundary Equilibrium Generative Adversarial Networks
• MatConvNet - Convolutional Neural Networks for MATLAB
• End-to-End Comparative Attention Networks for Person Re-identification
• Person Re-Identification by Camera Correlation Aware Feature Augmentation
• Deep Attributes Driven Multi-Camera Person Re-identification
• An Enhanced Deep Feature Representation for Person Re-identification
• Deep Ranking for Person Re-identification via Joint Representation Learning
• Unsupervised Person Re-identification: Clustering and Fine-tuning

Cited by in corpus (19)

• Person Re-identification by Contour Sketch under Moderate Clothing Change
• Interpretable and Generalizable Person Re-Identification with Query-Adaptive Convolution and Temporal Lifting
• A Self-Supervised Gait Encoding Approach with Locality-Awareness for 3D Skeleton Based Person Re-Identification
• Bi-directional Exponential Angular Triplet Loss for RGB-Infrared Person Re-Identification
• Learning to Adapt Invariance in Memory for Person Re-identification
• Hierarchical Skeleton Meta-Prototype Contrastive Learning with Hard Skeleton Mining for Unsupervised Person Re-Identification
• Survey on Reliable Deep Learning-Based Person Re-Identification Models: Are We There Yet?
• Exploring Shape Embedding for Cloth-Changing Person Re-Identification via 2D-3D Correspondences
• Memorizing Comprehensively to Learn Adaptively: Unsupervised Cross-Domain Person Re-ID with Multi-level Memory
• Global Distance-distributions Separation for Unsupervised Person Re-identification
• Person Re-ID through Unsupervised Hypergraph Rank Selection and Fusion
• Weakly Supervised Person Re-ID: Differentiable Graphical Learning and A New Benchmark
• Going Deeper into Semi-supervised Person Re-identification
• Intra-Camera Supervised Person Re-Identification
• Rethinking the Distribution Gap of Person Re-identification with Camera-based Batch Normalization
• Weakly Supervised Open-set Domain Adaptation by Dual-domain Collaboration
• Universal Person Re-Identification
• Weakly supervised discriminative feature learning with state information for person identification
• Semantic Consistency and Identity Mapping Multi-Component Generative Adversarial Network for Person Re-Identification