On Improving Deep Active Learning with Formal Verification
This work addresses data efficiency in neural-network training for machine learning practitioners, but it is incremental as it builds on existing deep active learning techniques.
The paper tackles the problem of reducing labeling costs in deep active learning by augmenting training data with adversarial inputs generated via formal verification, showing that this approach yields significant improvements in model generalization across standard benchmarks.
Deep Active Learning (DAL) aims to reduce labeling costs in neural-network training by prioritizing the most informative unlabeled samples for annotation. Beyond selecting which samples to label, several DAL approaches further enhance data efficiency by augmenting the training set with synthetic inputs that do not require additional manual labeling. In this work, we investigate how augmenting the training data with adversarial inputs that violate robustness constraints can improve DAL performance. We show that adversarial examples generated via formal verification contribute substantially more than those produced by standard, gradient-based attacks. We apply this extension to multiple modern DAL techniques, as well as to a new technique that we propose, and show that it yields significant improvements in model generalization across standard benchmarks.