Skip to content

Feelback: An Automated Video Feedback Application

License

Notifications You must be signed in to change notification settings

3omar-mostafa/Feelback

 
 

Repository files navigation

Feelback: An Automated Video Feedback Application

Feelback Logo

An Automated Video Feedback Application


Feelback is deployed at https://feelbackdemo.me Update: Azure hosting subscription has ended.


Video Demo

demo.mp4

Screenshots

video


overall


depth


keymoments


System Architecture

System Architecture



Install Dependencies

  • Install python >= 3.6
  • Install conda
  • conda env create -f requirements.yml

How to use

CLI

CLI is used only to create annotated video, visualizations, and testing

It DOES NOT generate analytics

Usage

Sample Usage

python feelback_cli.py input_video -o output -f native -v

All Options

Feelback is An Automated Video Feedback Framework

positional arguments:
  input_video           Input Video File Path

options:
  -h, --help            show this help message and exit
  -o filename, --output filename
                        Save processed video to filename.mp4
  --output-annotations annotations [annotations ...]
                        Which annotations to add to processed video
                        annotations can be any of: ['all', 'none', 'ids', 'age', 'gender', 'emotions', 'attention']
                        You can add multiple annotations separated by space
                        Default: all
                        
  --dump filename       Dump Feelback Object After Processing
  --load filename       Load Dumped Feelback Object [For Debug]
  --output-key-moments filename
                        Save Key Moments Visualizations to file
  -f N | native, --fps N | native
                        Process N frames every second, Or `native` to process all frames
  -v, --verbose         Enable more verbosity              

Backend

  • Create a .env file like the example.env file and change the following variables: UPLOAD_FOLDER, ANNOTATED_UPLOAD_FOLDER, DATABASE_URL, THUMBNAILS_FOLDER
  • run python feelback_server.py to run development server

Frontend

  • Install node.js >= 16
  • cd into feelback-front
  • run npm install
  • run npm run serve
  • to build for production, run npm run build

About

Feelback: An Automated Video Feedback Application

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 87.3%
  • Vue 10.9%
  • JavaScript 1.4%
  • Other 0.4%