InfractSizeDataset

• 

  Upload buttons for uploading files

• 

  Output of the dataset

• 

  Error message for uploading unequal files

Inspiration

This project is inspired by one of the computer vision machine learning models, Mask RCNN, I have trained.

What it does

This web program aims to facilitate the process of compressing images into a trainnable dataset format

How we built it

I build it using React as the frontend and Python Flask as the backend. Although sqlachemy is imported in this project, unfortunately I don't have time to implement it.

Challenges we ran into

It is my first time using React + Flask; therefore, I spend majority of my time setting up the environment and debugging for the API connections.

Accomplishments that we're proud of

Although this project is not completed (my end goal is to output predictions made by the ML model, but I am only able to finish the first part of creating trainable datasets), I am proud of my progress so far, particularly using a tech stack that I am first time using.

What we learned

I learned how React and Flask interact with each other! It is so different from React and Node.js, and of course it is giving me a lot of trouble debugging.

What's next for InfractSizeDataset

Next, I want to continually work on this project so that eventually it can be renamed to InfractSize—a website that, when you upload an infraction size image, predictions result on the affacted area will be display on the website.

Built With

• flask
• python
• react
• restapi