# DefectNET: multi-class fault detection on highly-imbalanced datasets

## TL;DR

DefectNET is a novel neural network architecture designed for multi-class fault detection in highly-imbalanced datasets, effectively detecting small and large defects with improved accuracy over existing methods.

## Contribution

The paper introduces DefectNet, combining parallel fully convolutional and dilated convolutional paths with a hybrid loss and leaky ReLU to enhance fault detection in imbalanced datasets.

## Key findings

- Outperforms state-of-the-art networks by ~10% accuracy on wind turbine data.
- Effectively detects small and large defects in highly-imbalanced datasets.
- Uses hybrid loss and leaky ReLU to improve training on rare defect classes.

## Abstract

As a data-driven method, the performance of deep convolutional neural networks (CNN) relies heavily on training data. The prediction results of traditional networks give a bias toward larger classes, which tend to be the background in the semantic segmentation task. This becomes a major problem for fault detection, where the targets appear very small on the images and vary in both types and sizes. In this paper we propose a new network architecture, DefectNet, that offers multi-class (including but not limited to) defect detection on highly-imbalanced datasets. DefectNet consists of two parallel paths, which are a fully convolutional network and a dilated convolutional network to detect large and small objects respectively. We propose a hybrid loss maximising the usefulness of a dice loss and a cross entropy loss, and we also employ the leaky rectified linear unit (ReLU) to deal with rare occurrence of some targets in training batches. The prediction results show that our DefectNet outperforms state-of-the-art networks for detecting multi-class defects with the average accuracy improvement of approximately 10% on a wind turbine.