Hard Samples Rectification for Unsupervised Cross-domain Person Re-identification
This work addresses the problem of person re-identification in unlabeled target domains for surveillance and security applications, representing an incremental improvement over existing clustering-based methods.
The paper tackles the challenge of unsupervised cross-domain person re-identification by proposing a Hard Samples Rectification learning scheme to address vulnerabilities to hard positive and negative samples, achieving promising results on large-scale benchmarks.
Person re-identification (re-ID) has received great success with the supervised learning methods. However, the task of unsupervised cross-domain re-ID is still challenging. In this paper, we propose a Hard Samples Rectification (HSR) learning scheme which resolves the weakness of original clustering-based methods being vulnerable to the hard positive and negative samples in the target unlabelled dataset. Our HSR contains two parts, an inter-camera mining method that helps recognize a person under different views (hard positive) and a part-based homogeneity technique that makes the model discriminate different persons but with similar appearance (hard negative). By rectifying those two hard cases, the re-ID model can learn effectively and achieve promising results on two large-scale benchmarks.