Let AI transform your images πŸ€–
Branch: master
Clone or download
Latest commit 3467062 Feb 1, 2019
Type Name Latest commit message Commit time
Failed to load latest commit information.
client initial commit Jan 30, 2019
demo initial commit Jan 30, 2019
server initial commit Jan 30, 2019
.gitignore initial commit Jan 30, 2019
LICENSE Create LICENSE Jan 31, 2019
README.md server typo Jan 30, 2019



Let AI transform your images

Installation: (run both client & server simultaneously)


cd client

# Install dependencies
npm install

# Start dev server
npm run serve


IMPORTANT: Download the model file on this link and put it in the server directory before doing the following.

cd server
pipenv shell

# Install dependencies (using Pipenv)
sudo pipenv install --skip-lock

# Start server with a category to detect
python main.py <category>

# Example:
python main.py person


You can start the server with one of these categories as arguments to detect them in your image.

  1. Vehicles:
  • car
  • bicycle
  • motorcycle
  • airplane
  • bus
  • train
  • truck
  • boat
  1. Animals:
  • bird
  • cat
  • dog
  • horse
  • sheep
  • cow
  • elephant
  • bear
  • zebra
  • giraffe
  1. Food:
  • banana
  • apple
  • orange
  • carrot
  • pizza
  • cake
  • cup
  • bottle
  1. Other:
  • person
  • umbrella
  • skateboard
  • laptop
  • tv
  • book


  • For some weird reason, you have to restart the server in order to transform another image. If you don't, you'll get the first image as output. I didn't have time to figure out why that happens.
  • You can't combine two categories.
  • By default, only the largest object will be detected.



Yeah. I know, I don't really deserve a donation for this tiny thing but here are my crypto addresses in case you're Jesus or just in case you want to support a 19 years old nerd.

  • Bitcoin: 1GUZ5w7A9PEu8TNrvKfAAaqhABUPW4RjfS

  • Ethereum: 0x56D6C69EC22Ee78A0454A165738A20a2e961385d

  • Bitcoin Cash: qz5mle4x2vjelax2c0wxhuprq5ha63w4xsj0r38g8s