Trustworthy AI-based crack-tip segmentation using domain-guided explanations

TL;DR

This paper introduces an attention-guided training framework that integrates explainability and domain knowledge to improve the trustworthiness and generalization of AI models in crack-tip segmentation tasks.

Contribution

It proposes a novel training approach that uses domain-specific explanations to guide model attention, enhancing physical relevance and robustness in high-stakes applications.

Findings

Improved model generalization in crack-tip segmentation.

Aligns model attention with physically meaningful stress fields.

Enhances trustworthiness through explainability and domain priors.

Abstract

Ensuring the trustworthiness and robustness of deep learning models remains a fundamental challenge, particularly in high-stakes scientific applications. In this study, we present a framework called attention-guided training that combines explainable artificial intelligence techniques with quantitative evaluation and domain-specific priors to guide model attention. We demonstrate that domain-specific feedback on model explanations during training can enhance the model's generalization capabilities. We validate our approach on the task of semantic crack tip segmentation in digital image correlation data, which is a key application in the fracture mechanical characterization of materials. By aligning model attention with physically meaningful stress fields, such as those described by Williams' analytical solution, attention-guided training ensures that the model focuses on physically relevant regions. This finally leads to improved generalization and more faithful explanations.