Abstract: Existing person re-identification (re-id) methods are stuck when deployed to a new unseen scenario despite the success in cross-camera person matching. Recent efforts have been substantially devoted to domain adaptive person re-id where extensive unlabeled data in the new scenario are utilized in a transductive learning manner. However, for each scenario, it is required to first collect enough data and then train such a domain adaptive re-id model, thus restricting their practical application. Instead, we aim to explore multiple labeled datasets to learn generalized domain-invariant representations for person re-id, which is expected universally effective for each new-coming re-id scenario. To pursue practicability in real-world systems, we collect all the person re-id datasets (20 datasets) in this field and select the three most frequently used datasets (i.e., Market1501, DukeMTMC, and MSMT17) as unseen target domains. In addition, we develop DataHunter that collects over 300K+ weak annotated images named YouTube-Human from YouTube street-view videos, which joins 17 remaining full labeled datasets to form multiple source domains. On such a large and challenging benchmark called FastHuman (~440K+ labeled images), we further propose a simple yet effective Semi-Supervised Knowledge Distillation (SSKD) framework. SSKD effectively exploits the weakly annotated data by assigning soft pseudo labels to YouTube-Human to improve models' generalization ability. Experiments on several protocols verify the effectiveness of the proposed SSKD framework on domain generalizable person re-id, which is even comparable to supervised learning on the target domains. Lastly, but most importantly, we hope the proposed benchmark FastHuman could bring the next development of domain generalizable person re-id algorithms.

Soft Pseudo-Label and Multi-Scale Feature Fusion for Person Re-Identification

DCLR-SF: Distribution Consistent Label Refinement and Lighten Similarity Network Fusion for Multi-Source Domain-Adaptive Person Re-Identification

Online Pseudo Label Generation by Hierarchical Cluster Dynamics for Adaptive Person Re-identification

Semi-Supervised Domain Generalizable Person Re-Identification

Group-aware Label Transfer for Domain Adaptive Person Re-identification

Pseudo-label Assisted Optimization of Multi-branch Network for Cross-domain Person Re-identification

Pseudo-Label Noise Prevention, Suppression and Softening for Unsupervised Person Re-Identification.

Domain-adaptive person re-identification via domain alignment and mutual pseudo-label refinement

Soft pseudo-Label shrinkage for unsupervised domain adaptive person re-identification

Learning Feature Fusion for Unsupervised Domain Adaptive Person Re-identification

A Person Re-Identification Method Based on Multi-Branch Feature Fusion

Refining Pseudo Labels for Unsupervised Domain Adaptive Person Re-Identification

Adaptive Style Normalization and Similarity Grouping for Unsupervised Cross-Domain Person Re-Identification

Multi-scale and Multi-Patch Feature Fusion Network for Person Re-Identification.

Feature Affinity-Based Pseudo Labeling for Semi-Supervised Person Re-Identification

Joint Cross-Consistency Learning and Multi-Feature Fusion for Person Re-Identification

A cross-domain person re-identification algorithm based on distribution-consistency and multi-label collaborative learning

A Novel Unsupervised Person Re-Identification Algorithm Based on Soft Multi-Label and Compound Attention Model

Unsupervised Person Re-Identification Based on High-Quality Pseudo Labels

Learning Feature Fusion for Unsupervised Domain Adaptive Person Re-identification

Complementary Pseudo Labels for Unsupervised Domain Adaptation on Person Re-identification.