Skip to content
Projects

Grad Courses

Road-Defect Classification with Deep Learning

Year
2025
Status
Graduate coursework
Category
Grad Courses
Tools
Python, PyTorch, ResNet18

Context

Explored computer vision for infrastructure condition assessment by classifying road-surface defects from images.

Approach

Applied transfer learning by fine-tuning a ResNet18 convolutional network on a road-surface defect dataset.

Outcome

Demonstrated a deep learning workflow for automated road-surface defect classification.