Labelfun

Labelfun is a web tool for computer vision annotation. It supports creating datasets for image classification, image object detection and video object detection.

On Labelfun, you can:

• Upload image or videos and create annotation tasks
• Label images and videos right in your browser
• Review new annotations
• Download annotations

Development

Server

1. Create server/.flaskenv with the content below:

   FLASK_APP=labelfun
   FLASK_ENV=development
   

2. Provide these environment variables in server/.env in the format of KEY=value:

   Key	Usage
   QINIU_ACCESS_KEY	The access key of your Qiniu Cloud account
   QINIU_SECRET_KEY	The secret key of your Qiniu Cloud account
   QINIU_BUCKET_DOMAIN	The domain name for you Qiniu Object Storage bucket
   QINIU_BUCKET_NAME	The name of your Qiniu Object Storage bucket
   INVITATION_CODE	The invitation code for a new account
   INVITATION_CODE_ADMIN	The invitation code for an admin account
   EXPORT_DIRECTORY	Where exported zip files are saved on the server
   LABELFUN_DATABASE_URL	The MySQL database URL (use mysqldb)

3. Under server/:

   $ python -m venv env
   
   > env\Scripts\activate  # Windows
   $ . env/bin/activate    # Linux / macOS
   
   (env) $ pip install -r requirements.txt
   (env) $ flask fakedb    # Initializes database
   (env) $ flask run

Client

1. Provide these environment variables in client/.env or client/.env.development or client/.env.production:

   Key	Usage
   VUE_APP_API	Where API requests are sent
   VUE_APP_QINIU_BUCKET_DOMAIN	The domain name for you Qiniu Object Storage bucket

2. Under client/:

   npm install
   npm run serve