# Teacher-Student Compression with Generative Adversarial Networks

arXiv: https://arxiv.org/abs/1812.02271 · HTML: https://mlsanity.com/paper/1812.02271/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 5 Dec 2018.

## Built on by (most cited first, 7 of 7)

- [Knowledge Distillation as Semiparametric Inference](https://mlsanity.com/paper/2104.09732/) (2021)
- [Categories of Response-Based, Feature-Based, and Relation-Based Knowledge Distillation](https://mlsanity.com/paper/2306.10687/) (2023)
- [Edge AI: Evaluation of Model Compression Techniques for Convolutional Neural Networks](https://mlsanity.com/paper/2409.02134/) (2024)
- [Efficient Encrypted Inference on Ensembles of Decision Trees](https://mlsanity.com/paper/2103.03411/) (2021)
- [A Scalable and Efficient Iterative Method for Copying Machine Learning Classifiers](https://mlsanity.com/paper/2302.02667/) (2023)
- [Logit Distillation on Manifolds: Mapping by Learning](https://mlsanity.com/paper/2606.00771/) (2026)
- [Distilling a Modular Reservoir Through a Genomic Bottleneck](https://mlsanity.com/paper/2606.28380/) (2026)

## Abstract (arXiv)

More accurate machine learning models often demand more computation and memory at test time, making them difficult to deploy on CPU- or memory-constrained devices. Teacher-student compression (TSC), also known as distillation, alleviates this burden by training a less expensive student model to mimic the expensive teacher model while maintaining most of the original accuracy. However, when fresh data is unavailable for the compression task, the teacher's training data is typically reused, leading to suboptimal compression. In this work, we propose to augment the compression dataset with synthetic data from a generative adversarial network (GAN) designed to approximate the training data distribution. Our GAN-assisted TSC (GAN-TSC) significantly improves student accuracy for expensive models such as large random forests and deep neural networks on both tabular and image datasets. Building on these results, we propose a comprehensive metric---the TSC Score---to evaluate the quality of synthetic datasets based on their induced TSC performance. The TSC Score captures both data diversity and class affinity, and we illustrate its benefits over the popular Inception Score in the context of image classification.