Efficient Semi-Supervised Adversarial Training via Latent Clustering-Based Data Reduction

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ghosh, Somrita, Xu, Yuelin, Zhang, Xiao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911494322520064
author Ghosh, Somrita
Xu, Yuelin
Zhang, Xiao
author_facet Ghosh, Somrita
Xu, Yuelin
Zhang, Xiao
contents Learning robust models under adversarial settings is widely recognized as requiring a considerably large number of training samples. Recent work proposes semi-supervised adversarial training (SSAT), which utilizes external unlabeled or synthetically generated data and is currently the state of the art. However, SSAT requires substantial extra data to attain high robustness, resulting in prolonged training time and increased memory usage. In this paper, we propose data reduction strategies to improve the efficiency of SSAT by optimizing the amount of additional data incorporated. Specifically, we design novel latent clustering-based techniques to select or generate a small, critical subset of data samples near the model's decision boundary. While focusing on boundary-adjacent points, our methods maintain a balanced ratio between boundary and non-boundary data points, thereby avoiding overfitting. Comprehensive experiments across image benchmarks demonstrate that our methods can effectively reduce SSAT's data requirements and computational costs while preserving its strong robustness advantages. In particular, our latent-space selection scheme based on k-means clustering and our guided diffusion-based approach with LCG-KM are the most effective, achieving nearly identical robust accuracies with 5 times to 10 times less unlabeled data. When compared to full SSAT trained to convergence, our methods reduce total runtime by approximately 3 times to 4 times due to strategic prioritization of unlabeled data.
format Preprint
id arxiv_https___arxiv_org_abs_2501_10466
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Efficient Semi-Supervised Adversarial Training via Latent Clustering-Based Data Reduction
Ghosh, Somrita
Xu, Yuelin
Zhang, Xiao
Machine Learning
Artificial Intelligence
Cryptography and Security
Computer Vision and Pattern Recognition
Learning robust models under adversarial settings is widely recognized as requiring a considerably large number of training samples. Recent work proposes semi-supervised adversarial training (SSAT), which utilizes external unlabeled or synthetically generated data and is currently the state of the art. However, SSAT requires substantial extra data to attain high robustness, resulting in prolonged training time and increased memory usage. In this paper, we propose data reduction strategies to improve the efficiency of SSAT by optimizing the amount of additional data incorporated. Specifically, we design novel latent clustering-based techniques to select or generate a small, critical subset of data samples near the model's decision boundary. While focusing on boundary-adjacent points, our methods maintain a balanced ratio between boundary and non-boundary data points, thereby avoiding overfitting. Comprehensive experiments across image benchmarks demonstrate that our methods can effectively reduce SSAT's data requirements and computational costs while preserving its strong robustness advantages. In particular, our latent-space selection scheme based on k-means clustering and our guided diffusion-based approach with LCG-KM are the most effective, achieving nearly identical robust accuracies with 5 times to 10 times less unlabeled data. When compared to full SSAT trained to convergence, our methods reduce total runtime by approximately 3 times to 4 times due to strategic prioritization of unlabeled data.
title Efficient Semi-Supervised Adversarial Training via Latent Clustering-Based Data Reduction
topic Machine Learning
Artificial Intelligence
Cryptography and Security
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2501.10466