aledelunap/fiberdiameter

Measure the diameter of fibers 10x faster

FiberDiameter

Measure the diameter of fibers in electron microscopy images, assisted by deep learning models, directly in your browser.

Usage

Online

The app is openly available at FiberDiameter.com

You can start by playing with the test image. First select the wand tool and then select the fiber that you want to measure.

The deep learning models are run directly in the browser. This means that your images never leave your computer and your intellectual property is preserved. In the world of machine learning, more data translates to better performance of the models. If you would like to donate images for training models, please get in touch with YaelSuarez. This will make the models better for you and for everyone.

Locally

git clone https://github.com/fcossio/FiberDiameter.git
cd FiberDiameter/frontend/
yarn # install node modules
yarn dev # start the development server

Contributing

Interested in contributing? Check out the contributing guidelines. Please note that this project is released with a Code of Conduct. By contributing to this project, you agree to abide by its terms.

License

FiberDiameter.com was created by Fernando Cossio, Yael Suarez, Luis Cossio and Ale de Luna. It is licensed under the terms of the Apache License 2.0 license.

Our tech stack

We use:

• kubrik to generate synthetic data.
• Tensorflow and Pytorch for training models (exported to ONNX).
• Nextjs for the frontend app.
• Pyodide and onnxruntime for running inference in the browser.
• And Vercel for deployment.

Aknowledgements

This project started as part of the W&B - Effective MLOps course and it was further developed during the Full Stack Deep Learning course.