Abstract:Ceramic substrates serve as the foundational material for numerous electronic devices, and their surface quality directly affects performance and longevity. Therefore, surface defect detection on ceramic substrates is an indispensable task during the manufacturing process. Ceramic substrate surface defects exhibit low contrast, along with intraclass difference and interclass similarity, posing substantial challenges for automated visual inspection. Existing deep learning-based methods utilize carefully crafted backbone networks and feature pyramid structures that perceive multilevel information to deal with the above challenges. However, these methods still suffer from insufficient extraction of discriminative features and low effectiveness in feature fusion, thereby hindering the detection performance. To address these issues, we propose the profound feature exploration and interaction network (PFEI-Net). Initially, we propose the discriminative feature mining backbone (DFM-backbone) to progressively explore the effective discriminative features. In the DFM-backbone, the global information grated perception module is devised to construct long-range feature dependencies and aggregate rich features, whereas the contextual semantic flexible extraction module is designed to meticulously perceive contextual semantic information in a targeted manner, enhancing the comprehensive representation of the discriminative features. Then, we propose the multipath semantic interaction guided-feature pyramid network (MSIG-FPN) to facilitate the fusion and interaction of the valuable information. In the MSIG-FPN, the hierarchical refocus–reaggregation module is constructed to refocus on task-relevant features and provide reliable guidance for the network, whereas the cross-stage semantic deep fusion module is designed to integrate multi-level contextual information deeply across stages and is employed to construct the semantic complementary paths, thereby preventing the loss of crucial features. Finally, built upon the backbone of DFM-backbone and MSIG-FPN, we establish the PFEI-Net, achieving accurate detection of surface defects on ceramic substrates. Experimental results highlight that the PFEI-Net achieves excellent performance with a remarkable 89.3 % mean average precision (mAP) and an inference time of 26.1 ms, which offers a promising and viable solution for automated surface defect detection of ceramic substrate.

DeePEB

PFEI-Net: A Profound Feature Exploration and Interaction Network for Ceramic Substrate Surface Defect Detection

DPO-SDF: Differentiable Physical Optics Method with Signed Distance Field for 3-D Reconstruction of Perfect Electric Conductors

Deeplab V3+ Based Segmentation Method for PV Panels with Aerial Orthoimages

Multi-step exposure method for improving structure flatness in digital light processing-based printing

LE-PDE++: Mamba for accelerating PDEs Simulations

Machine Learning Applied to Electron Beam Lithography to Accelerate Process Optimization of a Contact Hole Layer

Dynamic Exposure Model of Scanning Beam Interference Lithography

Machine learning in electron beam lithography to boost photoresist formulation design for high-resolution patterning

End-to-End Reaction Field Energy Modeling via Deep Learning based Voxel-to-voxel Transform

Feature Point-Based Proximity Effect Correction of Patterns

Layout Decomposition Co-Optimization for Hybrid E-Beam and Multiple Patterning Lithography.

Three-Dimensional Simulation of DRIE Process Based on the Narrow Band Level Set and Monte Carlo Method

Layout Decomposition For Hybrid E-Beam And Dsa Double Patterning Lithography

Programmable Photonic Extreme Learning Machines

Physics Informed Extreme Learning Machine (PIELM) -- A rapid method for the numerical solution of partial differential equations

Applying Machine Learning Models on Metrology Data for Predicting Device Electrical Performance

Canny Algorithm Enabling Precise Offline Line Edge Roughness Acquisition in High-Resolution Lithography.

Effective Optimization Strategy for Electron Beam Lithography of Molecular Glass Negative Photoresist

Groove Profile Prediction of Grating Masks in Scanning Beam Interference Lithography

A machine learning-based solution for modulated image analysis in lithography process stability diagnosis