Learning Incremental Triplet Margin for Person Re-identification

Abstract

Person re-identification (ReID) aims to match people across multiple non-overlapping video cameras deployed at different locations. To address this challenging problem, many metric learning approaches have been proposed, among which triplet loss is one of the state-of-the-arts. In this work, we explore the margin between positive and negative pairs of triplets and prove that large margin is beneficial. In particular, we propose a novel multi-stage training strategy which learns incremental triplet margin and improves triplet loss effectively. Multiple levels of feature maps are exploited to make the learned features more discriminative. Besides, we introduce global hard identity searching method to sample hard identities when generating a training batch. Extensive experiments on Market-1501, CUHK03, and DukeMTMCreID show that our approach yields a performance boost and outperforms most existing state-of-the-art methods.

accepted by AAAI19 as spotlight

References in corpus (8)

• Adam: A Method for Stochastic Optimization
• In Defense of the Triplet Loss for Person Re-Identification
• Random Erasing Data Augmentation
• Deep Representation Learning with Part Loss for Person Re-Identification
• AlignedReID: Surpassing Human-Level Performance in Person Re-Identification
• SphereReID: Deep Hypersphere Manifold Embedding for Person Re-Identification
• Beyond Part Models: Person Retrieval with Refined Part Pooling (and a Strong Convolutional Baseline)
• Deep-Person: Learning Discriminative Deep Features for Person Re-Identification