Skip to content

Latest commit

 

History

History
7 lines (4 loc) · 677 Bytes

README.md

File metadata and controls

7 lines (4 loc) · 677 Bytes

Attention U-Net

PyTorch implementation of "Attention U-Net: Learning Where to Look for the Pancreas." by Oktay et al applied for DRIVE blood vessels dataset. The model's implementation borrows from Hong Jing tutorial on Towards Data Science, available at: https://towardsdatascience.com/biomedical-image-segmentation-attention-u-net-29b6f0827405

The notebook describes the whole project process step by step, starting from describing the theoretical ideas that I've build my project upon, followed by implementing PyTorch Dataset class, the training loop and some visualizations.

The paper describing the architecture is available at: https://arxiv.org/pdf/1804.03999.pdf