Selfie Funhouse

NOTE THIS IS IN PROGRESS - YMMV

Example Output

Design

Here is the bigger picture of what this fits into:

About the code

It's Python2 - and it uses opencv+dlib.

Running

To run locally

Put some selfies into workdir/input and: python apply_mask.py -i workdir/input/ -o workdir/output/ -m funprops/

To build/run with AWS Lambda

You'll need to setup a few things in AWS as well as build a bundle.zip

I'm planning on leveraging the more generic work I'm doing here: https://github.com/dudash/aws-lambda-python-opencv TBD...

To add more props

You'll need to front (png) and back (jpg) files for each new prop into the funprops folder. They can be any size, but they must be the same size. The front is the prop and the back is how it should align to a face (any face will do).

---

The Computer Vision is 99% the awesome work from here: https://github.com/leoneckert/masks-and-hats

But I've modified it to (work on a Mac locally), be able to be run in AWS lambda, be able to use S3 for input and output, and added a bunch of props for the photos.