PyTorch implementation of UNet++ (Nested U-Net).

Last update: Jan 04, 2023

Overview

PyTorch implementation of UNet++ (Nested U-Net)

This repository contains code for a image segmentation model based on UNet++: A Nested U-Net Architecture for Medical Image Segmentation implemented in PyTorch.

[NEW] Add support for multi-class segmentation dataset.

[NEW] Add support for PyTorch 1.x.

Requirements

PyTorch 1.x or 0.41

Installation

Create an anaconda environment.

conda create -n=<env_name> python=3.6 anaconda
conda activate <env_name>

Install PyTorch.

conda install pytorch torchvision cudatoolkit=10.1 -c pytorch

Install pip packages.

pip install -r requirements.txt

Training on 2018 Data Science Bowl dataset

Download dataset from here to inputs/ and unzip. The file structure is the following:

inputs
└── data-science-bowl-2018
    ├── stage1_train
    |   ├── 00ae65...
    │   │   ├── images
    │   │   │   └── 00ae65...
    │   │   └── masks
    │   │       └── 00ae65...            
    │   ├── ...
    |
    ...

Preprocess.

python preprocess_dsb2018.py

Train the model.

python train.py --dataset dsb2018_96 --arch NestedUNet

Evaluate.

python val.py --name dsb2018_96_NestedUNet_woDS

(Optional) Using LovaszHingeLoss

Clone LovaszSoftmax from bermanmaxim/LovaszSoftmax.

git clone https://github.com/bermanmaxim/LovaszSoftmax.git

Train the model with LovaszHingeLoss.

python train.py --dataset dsb2018_96 --arch NestedUNet --loss LovaszHingeLoss

Training on original dataset

Make sure to put the files as the following structure (e.g. the number of classes is 2):

inputs
└── <dataset name>
    ├── images
    |   ├── 0a7e06.jpg
    │   ├── 0aab0a.jpg
    │   ├── 0b1761.jpg
    │   ├── ...
    |
    └── masks
        ├── 0
        |   ├── 0a7e06.png
        |   ├── 0aab0a.png
        |   ├── 0b1761.png
        |   ├── ...
        |
        └── 1
            ├── 0a7e06.png
            ├── 0aab0a.png
            ├── 0b1761.png
            ├── ...

Train the model.

python train.py --dataset <dataset name> --arch NestedUNet --img_ext .jpg --mask_ext .png

Evaluate.

python val.py --name <dataset name>_NestedUNet_woDS

Results

DSB2018 (96x96)

Here is the results on DSB2018 dataset (96x96) with LovaszHingeLoss.

Model	IoU	Loss
U-Net	0.839	0.365
Nested U-Net	0.842	0.354
Nested U-Net w/ Deepsupervision	0.843	0.362

PyTorch implementation of UNet++ (Nested U-Net).

Related tags

Overview

PyTorch implementation of UNet++ (Nested U-Net)

Requirements

Installation

Training on 2018 Data Science Bowl dataset

(Optional) Using LovaszHingeLoss

Training on original dataset

Results

DSB2018 (96x96)

Owner

4ui_iurz1

SAT Project - The first project I had done at General Assembly, performed EDA, data cleaning and created data visualizations

A solution to ensure Crowd Management with Contactless and Safe systems.

MODALS: Modality-agnostic Automated Data Augmentation in the Latent Space

Definition of a business problem according to Wilson Lower Bound Score and Time Based Average Rating

Face-Recognition-based-Attendance-System - An implementation of Attendance System in python.

An example of semantic segmentation using tensorflow in eager execution.

Python library for analysis of time series data including dimensionality reduction, clustering, and Markov model estimation

A FAIR dataset of TCV experimental results for validating edge/divertor turbulence models.

A collection of SOTA Image Classification Models in PyTorch

Breaching - Breaching privacy in federated learning scenarios for vision and text

OpenMMLab Computer Vision Foundation

Pytorch Lightning Distributed Accelerators using Ray

Alignment Attention Fusion framework for Few-Shot Object Detection

A Streamlit component to render ECharts.

A Real-Time-Strategy game for Deep Learning research

DNA-RECON { Automatic Web Reconnaissance Tool }

The source code for the Cutoff data augmentation approach proposed in this paper: "A Simple but Tough-to-Beat Data Augmentation Approach for Natural Language Understanding and Generation".

StyleGAN2-ada for practice

Unofficial pytorch implementation of 'Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization'

In this project we combine techniques from neural voice cloning and musical instrument synthesis to achieve good results from as little as 16 seconds of target data.