Nature

A multi-model deep learning framework for SEM-based defect detection in \(\textrm{FAPbI}_3\) Perovskite thin films

Integration of transfer learning and extensive data augmentation to overcome dataset scarcity. Benchmarking three architectures - ResNet50V2, DenseNet169 and YOLOv9 achieving 96.7% test accuracy.
Nature

A Public Image Dataset for Surface Defect Detection of Water-Based Coated Wood Products

Water-based wood coatings are environmentally friendly systems engineered specifically for wooden substrates. Characterized by the use of water as the primary solvent or dispersion medium, these ...
pv magazine International

Ensemble deep learning for PV cell defect detection

Researchers have tested eight stand-alone deep learning methods for PV cell fault detection and have found that their accuracy was as high as 73%. All methods were trained and tested on the ELPV ...
Semiconductor Engineering

Using Automatic Defect Classification To Reduce The Escape Rate Of Defects

Automated optical inspection (AOI) is a cornerstone in semiconductor manufacturing, assembly and testing facilities, and as such, it plays a crucial role in yield management and process control.
Semiconductor Engineering

Achieving Zero Defect Manufacturing Part 1: Detect & Classify

Whether the discussion is about smart manufacturing or digital transformation, one of the biggest conversations in the semiconductor industry today centers on the tremendous amount of data fabs ...