Attention Unet Pytorch

This repository contains an unofficial PyTorch implementation of the paper Attention U-Net: Learning Where to Look for the Pancreas by Zhu et al. The paper introduces an attention-based variation of the U-Net architecture for pancreas segmentation in medical imaging.

Paper: Attention U-Net: Learning Where to Look for the Pancreas

Overview

The Attention U-Net model is a modification of the U-Net architecture, incorporating self-attention mechanisms to selectively attend to informative regions in the image during both encoding and decoding. This approach enhances the model's ability to capture fine-grained details and spatial dependencies, leading to improved segmentation performance.

Dependencies

• Python 3.x
• PyTorch

to install all the dependencies run the following command:

pip3 install -r requirements.txt

Additional Information

This code is the implemented for the term project of CS 484 Introduction to Computer Vision at Bilkent University. The project is done by Yiğit Ekin and Arda Eren. The inference of this code on a dataset can be visualized by this link.