Track C1: Safety Verification of Deep Neural Networks (DNNs)

Formal verification of neural networks and broader machine learning models is an emerging field that has gained significant attention due to the growing use and impact of these data-driven methods. This track explores techniques for formally verifying neural networks and other machine learning models across various application domains. It includes papers and presentations discussing new methodologies, software frameworks, technical approaches, and case studies. Benchmarks play a crucial role in evaluating the effectiveness and scalability of these methods. Currently, available benchmarks mainly focus on computer vision problems, such as local robustness to adversarial perturbations of image classifiers. To address this limitation, this track compiles and publishes benchmarks comprising machine learning models and their specifications across domains such as computer vision, finance, security, and others. These benchmarks will help assess the suitability and applicability of formal verification methods in diverse domains.