Answers are grounded in this paper and cite their sources.
Abstract Detection of hidden corrosion within aircraft structures poses a significant challenge in nondestructive testing (NDT) methodologies, particularly in electromagnetic testing (ET) method. Corrosions often manifest in the recesses of rivet fastener holes, complicating their identification. In this paper, we propose an approach to enhance ET system detectability by introducing a Lightweight Magnetic Convolutional Neural Network (LMagNet) model tailored for the efficient processing of electromagnetic signals from corrosion. We aim to develop a lightweight model that could be deployed on low-resource microcontrollers ensuring its practical applicability in aircraft inspection scenarios. The proposed LMagNet model archives better performance compared to other lightweight models such as ShuffleNet, SqueezeNet, and MobileNet structures. The model archives an accuracy of 92% with only 16K parameters when being evaluated on hidden corrosion of aircraft structure, having volumes from 2.8-195.4 mm³. The model size is about 224×, 67×, 96× smaller compared to the ShuffleNet G1, SqueezeNet Complex, and MobileNet V3 model. When deploying on low-resource microcontrollers (i.e., STM32 MCUs), the LMagNet model requires only 90kB of flash and 36kB of RAM allowing to run within 40ms per prediction. In addition, we employed explainable techniques to interpret how the decision-making process of the model is made to achieve reliable results.
No takes yet. Share an insight, caveat, or question.
Vu et al. (2024) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: