Cluster-level Feature Alignment for Person Re-identification

15 Aug 2020  ·  Qiuyu Chen, Wei zhang, Jianping Fan ·

Instance-level alignment is widely exploited for person re-identification, e.g. spatial alignment, latent semantic alignment and triplet alignment. This paper probes another feature alignment modality, namely cluster-level feature alignment across whole dataset, where the model can see not only the sampled images in local mini-batch but the global feature distribution of the whole dataset from distilled anchors. Towards this aim, we propose anchor loss and investigate many variants of cluster-level feature alignment, which consists of iterative aggregation and alignment from the overview of dataset. Our extensive experiments have demonstrated that our methods can provide consistent and significant performance improvement with small training efforts after the saturation of traditional training. In both theoretical and experimental aspects, our proposed methods can result in more stable and guided optimization towards better representation and generalization for well-aligned embedding.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Person Re-Identification DukeMTMC-reID Cluster-level Alignment (Resnet50 w/o RK) Rank-1 91.11 # 21
mAP 81.84 # 30
Person Re-Identification Market-1501 Cluster-level Alignment (Resnet50 w/o RK) Rank-1 95.7 # 32
mAP 89.5 # 45


No methods listed for this paper. Add relevant methods here