Skip to content

Backend implementation that automates the annotation function of VGG Image Annotator(VIA) by using a trained neural network for inference.

License

Notifications You must be signed in to change notification settings

bkutasi/VIA-Auto

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Automated VGG Image Annotator (VIA)

Overview

This project is a backend implementation that automates the annotation function of VGG Image Annotator (VIA). It uses a trained neural network for inference, which significantly reduces the manual effort required in image annotation.

Features

  • Automated Annotation: The main feature of this project is to automate the process of image annotation. It uses a trained neural network model to predict annotations for a given image.
  • Integration with VIA: This project is designed and tested to work with VGG Image Annotator (VIA).
  • Specific domains: Currently only the heart slices dataset is supported shown in the demo.
  • Client side: VIA run entirely on the clients browser. This makes deployment very easily manageable with serverless architecture. An inference server is needed for VIA-Auto to work.

Getting Started

Please use Docker

docker pull tensorflow/tensorflow:latest-gpu-jupyter
docker run -it --rm -v $(realpath ~/notebooks):/tf/notebooks -p 8888:8888 tensorflow/tensorflow:latest-gpu-jupyter

To get started with this project, clone the repository and install the necessary dependencies.

Clone this repo

git clone https://github.com/bkutasi/VIA-Auto

Create an environment folder.

python3 -m venv env

Activate it

.\env\Scripts\activate

Install dependencies

pip install -r requirements.txt

Contributing

Contributions are welcome! Please read our contributing guidelines before starting.

Disclaimer

Since the original dataset seen in the demo was produced by SOTE researchers I wont release it. The documentation will contain everything you need to train your datasets besides bundled ones that are coming later.

Plans

  • Full client side: VIA-Auto to run entirely on the client with WASM Tensorflow.js integration thus becoming part of the VIA package.
  • VIA enhancements: VIA will be extended by some simple quality of life features that are deemed necessary.

License

This project is licensed under the terms of the GPLv3 license.

About

Backend implementation that automates the annotation function of VGG Image Annotator(VIA) by using a trained neural network for inference.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published