Image-AugmentationTP

Project description in one sentence

We decorate photos, videos but also a webcam live with global filters and face filters. ;)

Requirements

nodeJS

• only standart lib

python

• pandas
• torchfile
• Mediapipe
• numpy
• opencv
• pytorch(with cuda support)

How to Use

• Install node.js and python
• Install all Requirements (you can use the environment.yml, to install all requirements with our tested versions)
• Start main.py from Image-AugmentationTP ($python3 local_version/back/main.py)
• Start index.js from Image-AugmentationTP/local_version/front ($node index.js)
• (trust your server if you dont have https chrome://flags/#unsafely-treat-insecure-origin-as-secure)
• connect your client with the server (through the webbrowser)
• chose your favorite filter and click start stream :)

Concrete feature target list

• Use images as source for the filters (1.5)
• Use videos as source for the filters (8.5)
• Use webcam as life source for the filters (22.5) Not yet shown on the website
• Subpages for the different features (pictures videos webcam) (8.5)
• Program should recognize which filter to use when (15.5)
• face recognition filter (22.5)
• gangealing OR media pipe(5.06)
• style_transfer(5.06)
• Server-based approach (1 person at a time) / non-public (5.06)
• user controls program with buttons and dropdown menu / user friendly (19.06)