https://www.loom.com/share/cd07e6447bcc474a9257e1f0602bffa9

Navigate to /server folder, do a docker-compose up, and the backend is hosted and ready.

Project

• id : PK
• images : Image

Image

• id : PK
• url : link to local file directory
• annotations: Annotation[]

Annotation

• type
• startX
• startY
• endX
• endY

Navigate to /app folder, do a yarn install, and then yarn start to spin up the frontend.

1. User can create 'Project'

2. User can upload images to 'Project'

3. User can see images uploaded to 'Project'

4. User can do multiple annotations in seperate images in 'Project' (with 4 options)

5. User can download .csv file of image containing coordinations of bonding boxes of annotations (with added option?)

The task is to create a simple annotation framework. example 1::https://github.com/annotorious/annotorious example 2: https://youtu.be/ze0jXX4JEj4 The application should be server/client architecture Python Flask or Django, ReactJs preferred Step 1 : Add a feature where the user can create a project (Creates a directory using user input) upload 10 images related to traffic (Automobiles) Annotate 10 images On-screen a dropdown menu should be displayed with options car, bus, autorickshaw, bike Users should able to annotate multiple objects in an image. Save xy,xy,xy,xy coordinates of the individual image to a .csv file The format should image name and coordinates The manager should able view the annotated images with the bounding boxes Page writeup on how to start the application. Screenshots of the application for the corresponding task Example for .CSV file 71029695.jpg,930,594,1014,659 71029697.jpg,215,394,319,454 71029697.jpg,926,591,1019,658 Housekeeping rules: Do not share empty GitHub, GitLab or other code sharing platform links. Code should be uploaded to code sharing repo, with proper documentation, screenshots What do we look? Documentation Frontend knowledge Client-server architecture knowledge