# Embedding Expansion: Augmentation in Embedding Space for Deep Metric Learning

arXiv: https://arxiv.org/abs/2003.02546 · HTML: https://mlsanity.com/paper/2003.02546/

## Key facts

- Cited by 27 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Mar 2020.

## Built on by (most cited first, 27 of 27)

- [Improving Contrastive Learning by Visualizing Feature Transformation](https://mlsanity.com/paper/2108.02982/) (2021)
- [A Survey of Mix-based Data Augmentation: Taxonomy, Methods, Applications, and Explainability](https://mlsanity.com/paper/2212.10888/) (2022)
- [Human-Aided Saliency Maps Improve Generalization of Deep Learning](https://mlsanity.com/paper/2105.03492/) (2021)
- [It Takes Two to Tango: Mixup for Deep Metric Learning](https://mlsanity.com/paper/2106.04990/) (2021)
- [Embedding Transfer with Label Relaxation for Improved Metric Learning](https://mlsanity.com/paper/2103.14908/) (2021)
- [Learning with Memory-based Virtual Classes for Deep Metric Learning](https://mlsanity.com/paper/2103.16940/) (2021)
- [A Survey on Mixup Augmentations and Beyond](https://mlsanity.com/paper/2409.05202/) (2024)
- [Proxy Synthesis: Learning with Synthetic Classes for Deep Metric Learning](https://mlsanity.com/paper/2103.15454/) (2021)
- [Multi-Centroid Representation Network for Domain Adaptive Person Re-ID](https://mlsanity.com/paper/2112.11689/) (2021)
- [Granularity-aware Adaptation for Image Retrieval over Multiple Tasks](https://mlsanity.com/paper/2210.02254/) (2022)
- [Contrastive Bayesian Analysis for Deep Metric Learning](https://mlsanity.com/paper/2210.04402/) (2022)
- [LoOp: Looking for Optimal Hard Negative Embeddings for Deep Metric Learning](https://mlsanity.com/paper/2108.09335/) (2021)
- [Adaptive Feature Interpolation for Low-Shot Image Generation](https://mlsanity.com/paper/2112.02450/) (2021)
- [DAS: Densely-Anchored Sampling for Deep Metric Learning](https://mlsanity.com/paper/2208.00119/) (2022)
- [VVS: Video-to-Video Retrieval with Irrelevant Frame Suppression](https://mlsanity.com/paper/2303.08906/) (2023)
- [Globally Correlation-Aware Hard Negative Generation](https://mlsanity.com/paper/2411.13145/) (2024)
- [Two-Stage Triplet Loss Training with Curriculum Augmentation for Audio-Visual Retrieval](https://mlsanity.com/paper/2310.13451/) (2023)
- [Anchor-aware Deep Metric Learning for Audio-visual Retrieval](https://mlsanity.com/paper/2404.13789/) (2024)
- [Deep Metric Learning with Chance Constraints](https://mlsanity.com/paper/2209.09060/) (2022)
- [Construct Informative Triplet with Two-stage Hard-sample Generation](https://mlsanity.com/paper/2112.02259/) (2021)
- [Intra-class Adaptive Augmentation with Neighbor Correction for Deep Metric Learning](https://mlsanity.com/paper/2211.16264/) (2022)
- [Hotel Recognition via Latent Image Embedding](https://mlsanity.com/paper/2106.08042/) (2021)
- [Learning to Generate Novel Classes for Deep Metric Learning](https://mlsanity.com/paper/2201.01008/) (2022)
- [Generalized Sum Pooling for Metric Learning](https://mlsanity.com/paper/2308.09228/) (2023)
- [Multi-threshold Deep Metric Learning for Facial Expression Recognition](https://mlsanity.com/paper/2406.16434/) (2024)
- [Feature Augmentation for Self-supervised Contrastive Learning: A Closer Look](https://mlsanity.com/paper/2410.12396/) (2024)
- [OmniVaT: Single Domain Generalization for Multimodal Visual-Tactile Learning](https://mlsanity.com/paper/2601.00352/) (2026)

## Abstract (arXiv)

Learning the distance metric between pairs of samples has been studied for image retrieval and clustering. With the remarkable success of pair-based metric learning losses, recent works have proposed the use of generated synthetic points on metric learning losses for augmentation and generalization. However, these methods require additional generative networks along with the main network, which can lead to a larger model size, slower training speed, and harder optimization. Meanwhile, post-processing techniques, such as query expansion and database augmentation, have proposed the combination of feature points to obtain additional semantic information. In this paper, inspired by query expansion and database augmentation, we propose an augmentation method in an embedding space for pair-based metric learning losses, called embedding expansion. The proposed method generates synthetic points containing augmented information by a combination of feature points and performs hard negative pair mining to learn with the most informative feature representations. Because of its simplicity and flexibility, it can be used for existing metric learning losses without affecting model size, training speed, or optimization difficulty. Finally, the combination of embedding expansion and representative metric learning losses outperforms the state-of-the-art losses and previous sample generation methods in both image retrieval and clustering tasks. The implementation is publicly available.