# Learning Metrics from Teachers: Compact Networks for Image Embedding

Source: [Learning Metrics from Teachers: Compact Networks for Image Embedding](<https://www.sota2.com/research/paper/68fa65a01e209131cea0b5e7>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T21:31:37.819Z

Authors: Lu Yu, Vacit Oguz Yazici, Xialei Liu, Joost van de Weijer, Yongmei Cheng, Arnau Ramisa

Published: 2019-04-07T11:12:07.000Z

## Abstract

Metric learning networks are used to compute image embeddings, which are widely used in many applications such as image retrieval and face recognition. In this paper, we propose to use network distillation to efficiently compute image embeddings with small networks. Network distillation has been successfully applied to improve image classification, but has hardly been explored for metric learning. To do so, we propose two new loss functions that model the communication of a deep teacher network to a small student network. We evaluate our system in several datasets, including CUB-200-2011, Cars-196, Stanford Online Products and show that embeddings computed using small student networks perform significantly better than those computed using standard networks of similar size. Results on a very compact network (MobileNet-0.25), which can be used on mobile devices, show that the proposed method can greatly improve Recall@1 results from 27.5\\% to 44.6\\%. Furthermore, we investigate various aspects of distillation for embeddings, including hint and attention layers, semi-supervised learning and cross quality distillation. (Code is available at https://github.com/yulu0724/EmbeddingDistillation.)

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/1904.03624>)

[PDF](<https://arxiv.org/pdf/1904.03624.pdf>)
