Improving Performance of Semi-Supervised Learning by Adversarial Attacks
This work addresses the challenge of limited labeled data in SSL for image classification, but it appears incremental as it builds upon existing SSL algorithms.
The paper tackles the problem of improving semi-supervised learning (SSL) performance by using adversarial attacks to select high-confidence unlabeled data for labeling, resulting in significant improvements in image classification on CIFAR10.
Semi-supervised learning (SSL) algorithm is a setup built upon a realistic assumption that access to a large amount of labeled data is tough. In this study, we present a generalized framework, named SCAR, standing for Selecting Clean samples with Adversarial Robustness, for improving the performance of recent SSL algorithms. By adversarially attacking pre-trained models with semi-supervision, our framework shows substantial advances in classifying images. We introduce how adversarial attacks successfully select high-confident unlabeled data to be labeled with current predictions. On CIFAR10, three recent SSL algorithms with SCAR result in significantly improved image classification.