Mutual Mean-Teaching: Pseudo Label Refinery for Unsupervised Domain Adaptation on Person Re-identification

Abstract

Person re-identification (re-ID) aims at identifying the same persons' images across different cameras. However, domain diversities between different datasets pose an evident challenge for adapting the re-ID model trained on one dataset to another one. State-of-the-art unsupervised domain adaptation methods for person re-ID transferred the learned knowledge from the source domain by optimizing with pseudo labels created by clustering algorithms on the target domain. Although they achieved state-of-the-art performances, the inevitable label noise caused by the clustering procedure was ignored. Such noisy pseudo labels substantially hinders the model's capability on further improving feature representations on the target domain. In order to mitigate the effects of noisy pseudo labels, we propose to softly refine the pseudo labels in the target domain by proposing an unsupervised framework, Mutual Mean-Teaching (MMT), to learn better features from the target domain via off-line refined hard pseudo labels and on-line refined soft pseudo labels in an alternative training manner. In addition, the common practice is to adopt both the classification loss and the triplet loss jointly for achieving optimal performances in person re-ID models. However, conventional triplet loss cannot work with softly refined labels. To solve this problem, a novel soft softmax-triplet loss is proposed to support learning with soft pseudo triplet labels for achieving the optimal domain adaptation performance. The proposed MMT framework achieves considerable improvements of 14.4%, 18.2%, 13.1% and 16.4% mAP on Market-to-Duke, Duke-to-Market, Market-to-MSMT and Duke-to-MSMT unsupervised domain adaptation tasks. Code is available at https://github.com/yxgeee/MMT.

Accepted in International Conference on Learning Representations (ICLR 2020)

References in corpus (7)

• Distilling the Knowledge in a Neural Network
• Deep Domain Confusion: Maximizing for Domain Invariance
• Domain Separation Networks
• Adversarial Discriminative Domain Adaptation
• Unsupervised Person Re-identification by Soft Multilabel Learning
• Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation
• Invariance Matters: Exemplar Memory for Domain Adaptive Person Re-identification

Cited by in corpus (46)

• FastReID: A Pytorch Toolbox for General Instance Re-identification
• Complementary Pseudo Labels For Unsupervised Domain Adaptation On Person Re-identification
• Enhancing Diversity in Teacher-Student Networks via Asymmetric branches for Unsupervised Person Re-identification
• Robust Local Preserving and Global Aligning Network for Adversarial Domain Adaptation
• Test-time Batch Statistics Calibration for Covariate Shift
• Exploiting Sample Uncertainty for Domain Adaptive Person Re-Identification
• Unsupervised Two-Stage Anomaly Detection
• Multiple Expert Brainstorming for Domain Adaptive Person Re-identification
• Group-aware Label Transfer for Domain Adaptive Person Re-identification
• Joint Noise-Tolerant Learning and Meta Camera Shift Adaptation for Unsupervised Person Re-Identification
• Self-supervising Fine-grained Region Similarities for Large-scale Image Localization
• Generalizable Person Re-identification with Relevance-aware Mixture of Experts
• Intra-Inter Camera Similarity for Unsupervised Person Re-Identification
• Joint Generative and Contrastive Learning for Unsupervised Person Re-identification
• ST3D: Self-training for Unsupervised Domain Adaptation on 3D Object Detection
• Semantics-Guided Clustering with Deep Progressive Learning for Semi-Supervised Person Re-identification
• On Universal Black-Box Domain Adaptation
• Camera-aware Proxies for Unsupervised Person Re-Identification
• Re-identification = Retrieval + Verification: Back to Essence and Forward with a New Metric
• Improving Unsupervised Domain Adaptive Re-Identification via Source-Guided Selection of Pseudo-Labeling Hyperparameters
• Refining Pseudo Labels with Clustering Consensus over Generations for Unsupervised Object Re-identification
• IDM: An Intermediate Domain Module for Domain Adaptive Person Re-ID
• Improved Mutual Mean-Teaching for Unsupervised Domain Adaptive Re-ID
• On Evolving Attention Towards Domain Adaptation
• UnrealPerson: An Adaptive Pipeline towards Costless Person Re-identification
• Unsupervised Pre-training for Person Re-identification
• Towards Discriminative Representation Learning for Unsupervised Person Re-identification
• Disentanglement-based Cross-Domain Feature Augmentation for Effective Unsupervised Domain Adaptive Person Re-identification
• Discriminability Distillation in Group Representation Learning
• Layerwise Optimization by Gradient Decomposition for Continual Learning
• Graph Consistency Based Mean-Teaching for Unsupervised Domain Adaptive Person Re-Identification
• Deep Semi-supervised Knowledge Distillation for Overlapping Cervical Cell Instance Segmentation
• COCAS: A Large-Scale Clothes Changing Person Dataset for Re-identification
• Unsupervised Domain Adaptive Learning via Synthetic Data for Person Re-identification
• Multi-organ Segmentation via Co-training Weight-averaged Models from Few-organ Datasets
• Large-Scale Unsupervised Person Re-Identification with Contrastive Learning
• An Empirical Study of Vehicle Re-Identification on the AI City Challenge
• Complementary Relation Contrastive Distillation
• Take More Positives: An Empirical Study of Contrastive Learing in Unsupervised Person Re-Identification
• Unsupervised Person Re-identification via Simultaneous Clustering and Consistency Learning
• Cell Detection in Domain Shift Problem Using Pseudo-Cell-Position Heatmap
• Unsupervised Person Re-Identification with Multi-Label Learning Guided Self-Paced Clustering
• Deep High-Resolution Representation Learning for Cross-Resolution Person Re-identification
• Unsupervised Image Segmentation using Mutual Mean-Teaching
• Vision Meets Wireless Positioning: Effective Person Re-identification with Recurrent Context Propagation
• Adaptive Pseudo-Label Refinement by Negative Ensemble Learning for Source-Free Unsupervised Domain Adaptation