Enhancing robustness of data-driven SHM models: adversarial training with circle loss

Open in new window