Multi-level Distance Regularization For Deep Metric Learning

Abstract

We propose a novel distance-based regularization method for deep metric learning called Multi-level Distance Regularization (MDR). MDR explicitly disturbs a learning procedure by regularizing pairwise distances between embedding vectors into multiple levels that represents a degree of similarity between a pair. In the training stage, the model is trained with both MDR and an existing loss function of deep metric learning, simultaneously; the two losses interfere with the objective of each other, and it makes the learning process difficult. Moreover, MDR prevents some examples from being ignored or overly influenced in the learning process. These allow the parameters of the embedding network to be settle on a local optima with better generalization. Without bells and whistles, MDR with simple Triplet loss achieves the-state-of-the-art performance in various benchmark datasets: CUB-200-2011, Cars-196, Stanford Online Products, and In-Shop Clothes Retrieval. We extensively perform ablatio

Stats

• citations11
• S2 citations—
• github stars0
• HF likes0
• heat score8.09
• arxiv keykim2021multi

Related papers

• Cross-modal Deep Metric Learning With Multi-task Regularization (2017)9.03
• Deep Relational Metric Learning (2021)12.17
• Metric Learning With HORDE: High-order Regularizer For Deep Embeddings (2019)12.81
• Directional Statistics-based Deep Metric Learning For Image Classification And Retrieval (2018)13.05
• Guided Deep Metric Learning (2022)6.77
• Revisiting Training Strategies And Generalization Performance In Deep Metric Learning (2020)5.08
• Ranked List Loss For Deep Metric Learning (2019)7.50
• No Fuss Distance Metric Learning Using Proxies (2017)20.03