Reproduce “Learning Discriminative Prototypes via Boundary Repulsion MixUp for Object Re-Identification” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Athena Psalta, Vasileios Tsironis, Konstantinos Karantzalos
PBR-ReID trains learnable prototypes to find confusing pairs, then repels features from an interpolated proxy for the identities’ dividing region. At inference, it discards prototypes and retrieves using embeddings alone. In a Market-1501 ablation, it reaches 94.0% mAP, compared with 91.7% for neighbor-guided MixUp using the same pairs; this supports the operation’s role, but does not establish it as the sole cause.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.