Henry Alexander Lepp

Electrical engineering

Hometown: Las Vegas, Nevada, United States

Graduation date: Fall 2026

Additional details: Honors student

FURI | Spring 2025

Energy-efficient Wafer Defect Detection Using Spiking Neural Networks

Today’s approaches for silicon wafer defect detection involve Convolutional Neural Networks (CNNs) that result in high accuracy, at the cost of high computation and considerable power usage. One potential energy-efficient solution is to use Spiking Neural Networks (SNNs), which function like a brain, where neurons only “fire” at certain thresholds. Since neurons activate less often than in CNNs, SNNs may use fewer resources to train. The research team aims to train an SNN model that consumes less power with similar accuracy to CNN. This may enable engineers to implement energy-efficient defect detection with lower costs.

Mentor: Leslie Hwang