Unsupervised person re-identification with multi-label learning guided self-paced clustering

Although unsupervised person re-identification (Re-ID) has drawn increasing research attention recently, it remains challenging to learn discriminative features without annotations across disjoint camera views. In this paper, we address the Re-ID with a conceptually novel yet simple framework, termed as Multi-label Learning guided self-paced Clustering (MLC). MLC mainly learns three crucial modules, namely multi-scale network, multi-label learning module, and clustering module. Specifically, network generates multi-granularity in both global local The module leverages memory feature bank assigns each image vector based on similarities between bank. After training for several epochs, joins pseudo label image. benefits of our come from aspects: i) better similarity measurement, ii) assignment whole dataset ensures that every can be trained, iii) removes some noisy samples learning. Extensive experiments popular large-scale benchmarks demonstrate outperforms previous state-of-the-art methods significantly improves performance Re-ID.