Large-to-small Image Resolution Asymmetry In Deep Metric Learning

Abstract

Deep metric learning for vision is trained by optimizing a representation network to map (non-)matching image pairs to (non-)similar representations. During testing, which typically corresponds to image retrieval, both database and query examples are processed by the same network to obtain the representation used for similarity estimation and ranking. In this work, we explore an asymmetric setup by light-weight processing of the query at a small image resolution to enable fast representation extraction. The goal is to obtain a network for database examples that is trained to operate on large resolution images and benefits from fine-grained image details, and a second network for query examples that operates on small resolution images but preserves a representation space aligned with that of the database network. We achieve this with a distillation approach that transfers knowledge from a fixed teacher network to a student via a loss that operates per image and solely relies on coupled

Stats

• citations8
• S2 citations—
• github stars16
• HF likes0
• heat score9.62
• arxiv keysuma2022large

Related papers

• Three Things To Know About Deep Metric Learning (2024)0.00
• Asymmetric Metric Learning For Knowledge Transfer (2020)10.48
• Data-efficient Ranking Distillation For Image Retrieval (2020)0.00
• Learning Metrics From Teachers: Compact Networks For Image Embedding (2019)18.74
• Deep Metric Learning Using Similarities From Nonlinear Rank Approximations (2019)2.26
• Sharing Matters For Generalization In Deep Metric Learning (2020)8.35
• Dynamic Sampling For Deep Metric Learning (2020)5.84
• Visual Explanation For Deep Metric Learning (2019)14.36