Skip to content

Machine learning supported web-base image labelling tool for researcher.

License

Notifications You must be signed in to change notification settings

carpecodeum/LabelLab

 
 

Repository files navigation

LabelLab

License License

LabelLab is an image analyzing and classification platform. It allows users to upload batches of images and classify them with labels. It also has the features to run classifications against a trained model. LabelLab also has a user project management component as well as an image analyzing component.

Apache 2.0 Licence

See the LICENSE file for details.

Deployment

Host your own Labellab server for FREE with One-Click Deploy

Deploy

User Guide

How to Setup

Clone the repository.

git clone https://github.com/scorelab/Labellab.git

Set up React server

Run npm install in labellab-client folder.

 cd labellab-client/
 npm install

Set up MongoDB backend

Run npm install in labellab-server folder.

 cd labellab-server/
 npm install

How to Use

First you need to create a .env file in both labellab-server folder and labellab-client folder following the template provided in the file .env.example which is present in both the folders.

For client-side .env file:

REACT_APP_HOST=http://localhost
REACT_APP_SERVER_ENVIRONMENT=dev
REACT_APP_SERVER_PORT=4000

For server-side .env file:

HOST=localhost
PORT=4000

JWT_SECRET can be any word your choice.

To create a free MongoDB Atlas cloud database, follow the guide at https://docs.atlas.mongodb.com/getting-started/. After setting up, find the conection string to your database. It should look like this:

{{DB_HOST}}://{{DB_USERNAME}}:{{DB_PASSWORD}}@{{DB_CLUSTER}}/{{DB_NAME}}

You should fill in these values in their relevent fields in the .env file.

NOTE: Your IP address may change if you've been disconnected from the internet and reconnected. If that happens, it means that your internet connection doesn't have a static IP. You may no longer be able to access the database if you've whitelisted only your previous IP address. You can allow access from any IP address from the Atlas control panel to overcome this issue.

Both the front-end and the back-end can be run from the labellab-server folder using the terminal:

  1. To run both the client and server with a single command, run the following:
    $ npm run dev

  2. To run the server separately:
    $ npm run server

The server can also be run using $ npm start

  1. To run the client separately:
    $ npm run client

NOTE: Before starting the server create a file named .env same as .env.example and add your OAUTH and DATABASE credentials in the file.

Visit localhost:3000 to browse.

NOTE: This version only supports Google Chrome and Mozilla Firefox browsers. Make sure to instal the extension Redux Dev Tools if you're using Google Chrome.

How to Use (Mobile)

Run the Node server in the labellab-server directory (if not already done):

npm start

In a new terminal navigate to labellab_mobile directory.

Configure the lib/config.dart using the format provided in lib/config_example.dart.

The application needs google-services.json file issued from Firebase to operate Google Sign in option. To obtain this file,

  1. Sign in to https://console.firebase.google.com/.
  2. Click Add Project and necessary information about the project.
  3. Agree to the terms and click Create Project.
  4. After creating the project, click Add Firebase to your android app.
  5. Go to project location and open android/app/src/main/AndroidManifest.xml. Copy the package name(org.scorelab.labellab_mobile) and paste in the package name field
  6. Get the SHA1 fingerprint by following the instructions ( https://developers.google.com/android/guides/client-auth ) and paste it in the SHA1 fingerprint field.
  7. Click next and download the google-services.json.
  8. Paste the file in location android/app/ folder.

Build the Flutter application,

flutter build apk

NOTE: Use 'ios' instead of 'apk' to build for iOS.

or run the Flutter application.

flutter run

NOTE: A device with USB debugging enabled or virtual device is required.


For setting up the Flask backend

Technologies required

  • Python3 - Make sure you have a python3 environment setup.
  • Virtualenv - Make sure you install your packages locally in a virtual env and after adding a packing make sure to use
pip freeze>requirements.txt

  1. Make a database in your MYSQL

    CREATE {DATABASE | SCHEMA} [IF NOT EXISTS] labellab
  2. Create a virtual env in that folder using

    pip install virtualenv or sudo -H pip install virtualenv
    virtualenv -p python3 venv
  3. Change directory to the folder.

    cd venv/
    source bin/activate for MacOs or linux
    .\Source\activate.bat for Windows
  4. Change the directory back to the parent directory and
    add the credentials to your database in the .env file.

  5. Install all the dependencies using

    pip install -r requirements.txt
  6. Upgrade your database using the migrations file

    flask db upgrade
  7. Run flask run in order to start the server.

Note: It is preferred to undo the migrations before pulling the latest changes, if the latest changes have any changes in the migrations
or has a new migrations file and in order to create a new migartions file the command flask db migrate can be used.

Note: If faced by a directory missing issue do ensure that you have the labellab-flask directory in your PYTHONPATH environment variable.

About

Machine learning supported web-base image labelling tool for researcher.

Resources

License

Code of conduct

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • JavaScript 54.5%
  • Dart 31.1%
  • Python 9.2%
  • CSS 2.7%
  • TSQL 1.3%
  • Ruby 0.6%
  • Other 0.6%