Lie Group Intrinsic Mean Feature Detectors: An Effective Surface Defect Detection Model for Real-Time Industrial Applications

Abstract

With the rapid popularization of the Internet of Things (IoT) and the widespread application of deep learning, more and more IoT devices are embedded with artificial intelligence (AI) processors, which can achieve efficient edge detection processing, such as surface defect detection (SSD). However, in the actual industrial production environment, the surface defects of products are quite tiny, and the number of different types of defect data samples is also quite small, most deep learning models rely on a large number of training samples and parameters to achieve high-precision defect detection. At the same time, the edge computing layer in the actual industrial environment may also encounter transmission delay and insufficient resources. Training a proper model for a specific type of surface defect and simultaneously satisfying the real-time and accuracy of defect detection is still a challenging task. To effectively deal with the above challenges, we propose an edge-cloud computing defect detection model based on the intrinsic mean feature detector in the Lie Group manifold space, which utilizes the Lie Group manifold space intrinsic mean feature as a metric to characterize the essential attributes of different types of surface defects. In addition, we propose an intrinsic mean attention mechanism in the Lie Group manifold space that is easy to implement at the edge service layer without increasing the number of model parameters, enhancing the detection performance of tiny surface defects. Extensive experiments on three publicly available and challenging datasets reveal the superiority of our model in terms of detection accuracy, detection real-time, number of parameters, and computational performance. In addition, our proposed model also shows some competitiveness and advantages compared with the state-of-the-art models.

Keywords: Attention mechanism, Lie Group, Multi-scale feature extraction and fusion, surface defect detection