Dynamic Sampling For Deep Metric Learning

Abstract

Deep metric learning maps visually similar images onto nearby locations and visually dissimilar images apart from each other in an embedding manifold. The learning process is mainly based on the supplied image negative and positive training pairs. In this paper, a dynamic sampling strategy is proposed to organize the training pairs in an easy-to-hard order to feed into the network. It allows the network to learn general boundaries between categories from the easy training pairs at its early stages and finalize the details of the model mainly relying on the hard training samples in the later. Compared to the existing training sample mining approaches, the hard samples are mined with little harm to the learned general model. This dynamic sampling strategy is formularized as two simple terms that are compatible with various loss functions. Consistent performance boost is observed when it is integrated with several popular loss functions on fashion search, fine-grained classification, and

Stats

• citations5
• S2 citations—
• github stars0
• HF likes0
• heat score5.84
• arxiv keyliang2020dynamic

Related papers

• Symmetrical Synthesis For Deep Metric Learning (2020)8.60
• Graph Sampling Based Deep Metric Learning For Generalizable Person Re-identification (2021)20.14
• Guided Deep Metric Learning (2022)6.77
• Sharing Matters For Generalization In Deep Metric Learning (2020)8.35
• Deep Metric Learning For Computer Vision: A Brief Overview (2023)6.77
• The Group Loss For Deep Metric Learning (2019)11.39
• Loop: Looking For Optimal Hard Negative Embeddings For Deep Metric Learning (2021)8.82
• DAS: Densely-anchored Sampling For Deep Metric Learning (2022)9.76