All You Need to Know About Training Image Retrieval Models
Not on the map. Why?
Abstract (arXiv)
Image retrieval is the task of finding images in a database that are most similar to a given query image. The performance of an image retrieval pipeline depends on many training-time factors, including the embedding model architecture, loss function, data sampler, mining function, learning rate(s), and batch size. In this work, we run tens of thousands of training runs to understand the effect each of these factors has on retrieval accuracy. We also discover best practices that hold across multiple datasets. The code is available at https://github.com/gmberton/image-retrieval
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 14 of its 24 references that are in the map. First posted on arXiv 17 Mar 2025.
Builds on
14 of 24 references.
References · 14
- 2019RandAugment: Practical automated data augmentation with a reduced search spacecited by 2,261
- 2018CosFace: Large Margin Cosine Loss for Deep Face Recognitioncited by 968
- 2015Deep Metric Learning via Lifted Structured Feature Embeddingcited by 449
- 2020Circle Loss: A Unified Perspective of Pair Similarity Optimizationcited by 402
- 2019Multi-Similarity Loss with General Pair Weighting for Deep Metric Learningcited by 368
- 2020A Metric Learning Reality Checkcited by 317
- 2020Proxy Anchor Loss for Deep Metric Learningcited by 195
- 2019SoftTriple Loss: Deep Metric Learning Without Triplet Samplingcited by 163
- 2020Revisiting Training Strategies and Generalization Performance in Deep Metric Learningcited by 99
- 2019Signal-to-Noise Ratio: A Robust Distance Metric for Deep Metric Learningcited by 32
- 2021Characterizing Generalization under Out-Of-Distribution Shifts in Deep Metric Learningcited by 17
- 2019Unbiased Evaluation of Deep Metric Learning Algorithmscited by 10
Built on by
Citing papers per year: 2026: 1.
1 paper.