wafer-defects

We use MixedWM38, the mixed-type wafer defect pattern dataset for wafer defect pattern regcognition with visual transformers.

End to end machine learning project for accurately detecting faults in wafers based on sensor data.

This project aims to process 2D images of semiconductor silicon wafers to identify any defects on the wafers as well as their corresponding locations.

Classification of wafer defect map patterns

Official implementation of "Hybrid Quantum-MambaVision" paper accepted at PAKDD 2026.

This is a neural network designed to classify wafer imperfections without feature engineering.

Drift-Sense solves the Navigation-Error Recovery problem. By utilizing a hybrid architecture of classical vision (NCC) and a lightweight Siamese CNN ranker, it accurately matches high-resolution reference patterns against noisy, low-resolution SEM images without heavy transformers.

Identify silicon wafers by their unique edge-blob fingerprint using DISK + LightGlue

data fetched by wafers (thin slices of semiconductors) is to be passed through the machine learning pipeline and it is to be determined whether the wafer at hand is faulty or not. Wafers are predominantly used to manufacture solar cells and are located at remote locations in bulk and they themselves consist of few hundreds of sensors.

Lightweight YOLOv11 for real-time micron wafer defect industrial inspection with TensorRT deployment

High-fidelity process simulation dashboard built in Python. Features include Deal-Grove thermal oxidation (calibrated to BYU standards), Gaussian ion implantation profiling, and selective etch rate modeling with real-time cross-sectional visualization.

A reproducible deep-learning study on semiconductor wafer defect classification (WM-811K), comparing custom CNNs, model compression, and transfer learning on imbalanced, discrete-value data.

清纯AOI优化与叠图

End-to-end deep learning app for detecting defects in semiconductor wafermaps

Wafer Defect Pattern Classification project

This repo contains well-designed tabular features created through image processing techniques in order to predict a scratch on wafers.

Faulty Wafer Detection

data fetched by wafers (thin slices of semiconductors) is to be passed through the machine learning pipeline and it is to be determined whether the wafer at hand is faulty or not. Wafers are predominantly used to manufacture solar cells and are located at remote locations in bulk and they themselves consist of few hundreds of sensors.

This repository hosts a comprehensive end-to-end machine learning project focused on Wafer Fault Detection, implemented using Python with Flask.