Generating artificial digital image correlation data using physics-guided adversarial networks
This addresses data scarcity for crack detection in mechanical experiments, though it is incremental as it builds on existing GAN approaches with added physics guidance.
The paper tackled the problem of scarce training data for crack detection in digital image correlation by generating artificial displacement data using a physics-guided GAN, which improved visual quality and metrics like sliced Wasserstein distance compared to unguided methods.
Digital image correlation (DIC) has become a valuable tool to monitor and evaluate mechanical experiments of cracked specimen, but the automatic detection of cracks is often difficult due to inherent noise and artefacts. Machine learning models have been extremely successful in detecting crack paths and crack tips using DIC-measured, interpolated full-field displacements as input to a convolution-based segmentation model. Still, big data is needed to train such models. However, scientific data is often scarce as experiments are expensive and time-consuming. In this work, we present a method to directly generate large amounts of artificial displacement data of cracked specimen resembling real interpolated DIC displacements. The approach is based on generative adversarial networks (GANs). During training, the discriminator receives physical domain knowledge in the form of the derived von Mises equivalent strain. We show that this physics-guided approach leads to improved results in terms of visual quality of samples, sliced Wasserstein distance, and geometry score when compared to a classical unguided GAN approach.