# FoQuS: A Forgetting-Quality Coreset Selection Framework for Automatic Modulation Recognition

arXiv: https://arxiv.org/abs/2509.08300 · HTML: https://mlsanity.com/paper/2509.08300/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 21 references that are in the map.
- First posted on arXiv 10 Sep 2025.

## Builds on (most cited first, 13 of 13)

- [An Empirical Study of Example Forgetting during Deep Neural Network Learning](https://mlsanity.com/paper/1812.05159/) (2018)
- [Deep Learning on a Data Diet: Finding Important Examples Early in Training](https://mlsanity.com/paper/2107.07075/) (2021)
- [Selection via Proxy: Efficient Data Selection for Deep Learning](https://mlsanity.com/paper/1906.11829/) (2019)
- [Active Learning by Acquiring Contrastive Examples](https://mlsanity.com/paper/2109.03764/) (2021)
- [DeepCore: A Comprehensive Library for Coreset Selection in Deep Learning](https://mlsanity.com/paper/2204.08499/) (2022)
- [Contextual Diversity for Active Learning](https://mlsanity.com/paper/2008.05723/) (2020)
- [Over the Air Deep Learning Based Radio Signal Classification](https://mlsanity.com/paper/1712.04578/) (2017)
- [Online Coreset Selection for Rehearsal-based Continual Learning](https://mlsanity.com/paper/2106.01085/) (2021)
- [Coreset Selection for Object Detection](https://mlsanity.com/paper/2404.09161/) (2024)
- [SigNet: A Novel Deep Learning Framework for Radio Signal Classification](https://mlsanity.com/paper/2011.03525/) (2020)
- [Dataset Distillation Meets Provable Subset Selection](https://mlsanity.com/paper/2307.08086/) (2023)
- [RK-core: An Established Methodology for Exploring the Hierarchical Structure within Datasets](https://mlsanity.com/paper/2310.12168/) (2023)
- [Non-Uniform Class-Wise Coreset Selection for Vision Model Fine-tuning](https://mlsanity.com/paper/2504.13234/) (2025)

## Abstract (arXiv)

Deep learning-based Automatic Modulation Recognition (AMR) model has made significant progress with the support of large-scale labeled data. However, when developing new models or performing hyperparameter tuning, the time and energy consumption associated with repeated training using massive amounts of data are often unbearable. To address the above challenges, we propose FoQuS, which approximates the effect of full training by selecting a coreset from the original dataset, thereby significantly reducing training overhead. Specifically, FoQuS records the prediction trajectory of each sample during full-dataset training and constructs three importance metrics based on training dynamics. Experiments show that FoQuS can maintain high recognition accuracy and good cross-architecture generalization on multiple AMR datasets using only 1%-30% of the original data.