Cracks Segmentation
Hamzah ElQadasi
AI Model Developer
PyTorch
scikit-learn
seaborn
University of Houston-Data Science Institute
In my Crack Segmentation project, in partnership with Houston University's structural engineering team, I utilized advanced machine learning techniques to significantly enhance crack detection in images. Through the integration of tools like UNet++, EfficientNet-B3, and PyTorch, I increased pixel detection accuracy from 70% to 86%, improving the effectiveness of damage assessment in structural health monitoring.
